II — The Mechanisms
Chapter 8: The Architecture of Secrecy
The preceding chapter documented the silence — the simultaneous cessation of public gravitics research across an entire industry. The question it raised was how such a silence could be produced and sustained. The answer lies in a legal and institutional architecture whose components are, individually, matters of public record.
The system is not hidden. Its statutes are published in the United States Code. Its executive orders are printed in the Federal Register. Its accounting standards are posted on government websites. Its directives are issued by the Department of Defence. Its court decisions are reported in the federal reporters. Every component of the architecture described in this chapter is a matter of public record. The architecture of secrecy operates in the open. Its applications are what disappear.
I. The Invention Secrecy Act
The legal foundation of technology suppression in the United States is not an executive order, not a classified directive, not a secret agreement between intelligence agencies. It is a statute, passed by Congress, signed by the President, codified at Title 35, United States Code, Sections 181 through 188. It is called the Invention Secrecy Act, and it has been in continuous operation since 1 February 1952.
The Act's predecessor was Public Law 700 of the 77th Congress, enacted in 1941 as a wartime measure giving the Commissioner of Patents authority to keep secret any patent application disclosing inventions useful to national defence. That authority expired six months after the war's end. The Cold War made the authority permanent. The Korean War served as the proximate catalyst. The secrecy provisions were embedded in a much larger patent reform bill -- the Patent Act of 1952, Public Law 82-256 -- and did not receive standalone scrutiny. Floor debate was minimal. The provisions drew on recommendations from the Department of Defence and the Atomic Energy Commission and passed with bipartisan support.
The mechanism operates step by step, and each step deserves attention.
An inventor files a patent application with the United States Patent and Trademark Office. This is the normal path for establishing intellectual property rights -- the mechanism by which an inventor protects a discovery and brings it to market. It is the first step toward commercialisation. And it is the point of interception.
The USPTO maintains an internal programme, managed by its Licensing and Review Branch, that screens incoming patent applications against criteria established by defence and intelligence agencies. The screening criteria are contained in a document called the Patent Security Category Review List -- the PSCRL -- maintained by the Armed Services Patent Advisory Board. The PSCRL is itself classified, or at minimum restricted. Its contents are not publicly available. But a version from approximately 1971 became known through researchers and is referenced in Federation of American Scientists publications. Among the categories flagged for secrecy review in that version:
Solar photovoltaic cells with conversion efficiency exceeding twenty per cent. Energy conversion systems exceeding seventy to eighty per cent efficiency. Advanced propulsion concepts. Weapons-related categories including guided missiles, nuclear technology, cryptography, stealth and radar absorption, and underwater sound propagation. Certain materials science categories.
The solar cell threshold demands attention. In 1971, commercially available solar cells were well below twenty per cent efficiency. A secrecy order on patents exceeding that threshold would directly retard civilian energy technology. This is not a weapons category. It is an energy category. The energy conversion efficiency threshold demands equal attention: seventy to eighty per cent exceeds the theoretical limits of most conventional heat engines, meaning that any genuinely novel energy conversion method -- any technology that approached the thermodynamic ideal -- would be flagged. The PSCRL demonstrates that the screening criteria extend beyond weapons to encompass technologies that would disrupt the existing energy infrastructure -- precisely the category into which ether-based energy technology falls.
The PSCRL has been updated multiple times since 1971. Later versions are not public. The categories almost certainly now include quantum computing, directed energy, hypersonic technologies, and space-related technologies. What the list contained in 1956, when the gravitics programmes went silent, or in 1943, when Tesla's papers were seized, is unknown. The list is classified. But the 1971 version reveals the principle: the screening net is cast wider than weapons. It captures energy. It captures efficiency. It captures propulsion. The categories that would encompass ether-based technology -- novel energy conversion, advanced propulsion, exotic materials -- are not hypothetical additions to the PSCRL. They are the categories it was designed to capture.
If an application is flagged during screening, it is routed to the relevant defence agency for review. The agencies with review authority include the Department of Defence (often through the Defence Technical Information Centre or service-specific entities), the Department of Energy (successor to the Atomic Energy Commission for nuclear-related matters), the National Security Agency (for cryptography and signals intelligence), the Department of Justice and FBI, NASA, and any other agency designated by the President. The reviewing agency determines whether disclosure of the invention could be "detrimental to the national security." There is no statutory standard. There is no burden of proof. The determination is discretionary. The statute uses the phrase "in the opinion of" the relevant agency head -- a formulation that grants maximum latitude and minimum reviewability.
If the agency recommends secrecy, the USPTO Director issues a Secrecy Order. The inventor receives a letter -- typically by registered mail -- notifying them that the application is under a secrecy order, that they may not disclose the invention to any unauthorised person, that they may not file foreign patent applications, and that violation is a federal crime. Section 186 of the Act specifies the penalty: a fine of not more than ten thousand dollars, imprisonment for not more than two years, or both. Section 187 exempts officers and employees of the United States acting within the scope of their duties -- a provision whose asymmetry speaks for itself: the government's servants are exempt from the penalties that bind the private citizen whose invention the government has seized.
The patent application is not abandoned. It continues to be examined. But no patent will be granted, and no application will be published, for as long as the secrecy order remains in effect. The inventor's priority date is preserved -- a nominal concession, given that the priority date is meaningless if the patent is never issued. The order is reviewed annually. The relevant agency can renew it. There is no limit on the number of renewals. The Federation of American Scientists has documented cases of orders remaining in effect for thirty, forty, and fifty years or more. There is no sunset provision. An order remains until the relevant agency determines the invention no longer needs protection, which may be never. As Steven Aftergood of FAS has noted, the lack of a mandatory expiration or mandatory periodic review by an independent body means that secrecy orders can outlast the original national security rationale. A technology classified in 1960 because it might advantage the Soviet Union can remain classified in 2025 -- sixty-five years later, thirty-four years after the Soviet Union ceased to exist.
Section 183 provides a theoretical right to compensation. The inventor may apply to the agency that caused the order for damages. If denied, the inventor may sue in the Court of Federal Claims. The mechanism looks reasonable on paper. In practice, it is designed to fail. The inventor cannot freely discuss the invention -- making valuation impossible without exposing classified details. Legal representation is complicated by the inability to disclose those details to counsel without the counsel holding the appropriate clearance. The agency that caused the secrecy order is the same agency that determines whether compensation is owed -- a conflict of interest embedded in the statute. The government has little incentive to offer generous settlements; each case is isolated and invisible, creating no public pressure. Few inventors have the resources to litigate against the federal government in the Court of Federal Claims, a specialised tribunal where the government is always the defendant. There is no public accounting of total compensation paid under this section. One of the few cases to reach the courts, Halpern v. United States (Court of Claims, 1958), demonstrated the difficulty: the court acknowledged the inventor's right to compensation but wrestled with how to value an invention that the inventor could not describe and the public could not examine. The compensation mechanism exists on paper. In practice, it is an obstacle course designed for the government to win.
The most restrictive form of secrecy order is the Type III General Secrecy Order, which prohibits all disclosure, all publication, and all foreign filing. No exceptions. No permits. The less restrictive forms -- Type I, which permits some foreign filing in certain countries under conditions, and Type II, with intermediate restrictions -- at least acknowledge that an inventor has residual rights. Type III acknowledges nothing. It is total suppression.
First Amendment scholars recognise the Type III order as a prior restraint on speech -- one of the most disfavoured forms of government restriction under constitutional law. Prior restraints bear a "heavy presumption" of unconstitutionality under the Supreme Court's holdings in Bantam Books v. Sullivan (1963) and New York Times Co. v. United States (1971, the Pentagon Papers case). The Invention Secrecy Act has survived constitutional challenge on the theory that patent applications are a special regulatory context and that national security justifies the restraint. No court has squarely held the Act unconstitutional. The constitutional question has never been fully litigated to the Supreme Court. The Act operates in a zone of low judicial scrutiny because national security deference is strong and affected parties are isolated from each other and from public support.
The Fifth Amendment implications are equally serious. A secrecy order effectively takes the inventor's property right -- the patent monopoly, which has tangible commercial value -- without compensation for the duration of the order. The theoretical compensation mechanism under Section 183 does not change the structural reality: the government takes the property first and litigates compensation later, if the inventor can afford to fight. Legal scholars have argued that this constitutes an uncompensated taking in violation of the Fifth Amendment's Takings Clause. The argument has never prevailed because no court has been willing to hold the Act unconstitutional in the face of national security claims. The legal challenges that have been mounted are vanishingly few -- civil liberties organisations including the ACLU and EFF have raised concerns, but no direct facial challenge has succeeded, partly because standing is nearly impossible to establish. The affected inventors are gagged. They cannot publicise their situation without risking prosecution. They cannot organise because they do not know who the other affected inventors are.
Here is the fact that elevates the Invention Secrecy Act from a narrowly technical provision to a structural instrument of suppression: the Act applies to purely private inventions. Section 181 extends explicitly to inventions "in which the Government does not have a property interest." No government contract is required. No government funding is required. A private citizen, working in a private laboratory, with private money, who files a patent on a device that the defence establishment considers sensitive, can be silenced under criminal penalty, indefinitely, with no meaningful recourse. The statute's own language is worth reading in full on this point: "whenever the publication or disclosure of an invention by the publication of an application or by the granting of a patent, in which the Government does not have a property interest, might, in the opinion of the Commissioner of Patents, be detrimental to the national security, he shall make the application for patent in which such invention is disclosed available for inspection" to the relevant defence agencies. The reach is explicit. The constraint is absolute.
The international dimension compounds the suppression. Section 184 prohibits filing a patent application in any foreign country for any invention made in the United States without first obtaining a foreign filing licence from the USPTO Director. For inventions under secrecy orders, no such licence will be granted. This interacts with the Patent Cooperation Treaty, administered by the World Intellectual Property Organisation, which normally allows a single international filing to designate multiple countries. If the invention is made in the United States and is under a secrecy order, the inventor cannot file a PCT application. The national security override takes precedence over international treaty obligations. The practical effect: an inventor under a secrecy order loses not only the United States patent but all foreign patent rights for the duration of the order. If the order persists for decades and is then lifted, the technology may be obsolete and the patent term substantially consumed. The inventor loses not just the domestic market but the global one. Other nations have equivalent mechanisms -- the United Kingdom's Section 22 of the Patents Act 1977, France's Articles L. 612-9 and L. 612-10 of the Intellectual Property Code -- but the United States system is the most extensive in absolute numbers and the broadest in its application to private inventions.
The statistics tell their own story. The Federation of American Scientists, primarily through Steven Aftergood's Secrecy News project, has tracked the data reported by the USPTO to Congress. As of fiscal year 2025, approximately 6,543 secrecy orders are active. The annual new order figures across selected years paint the trend: approximately 122 in FY2010, 103 in FY2011, 120 in FY2012, 128 in FY2013, 101 in FY2014, 120 in FY2015, 97 in FY2016, 143 in FY2017, 109 in FY2018, 110 in FY2019, 73 in FY2020, 98 in FY2021, 107 in FY2022, 125 in FY2023. The long-term range for new orders was typically seventy to one hundred and fifty per year through the 2000s and 2010s. In fiscal year 2024, 356 new secrecy orders were imposed -- nearly triple the typical annual rate. Aftergood flagged this in Secrecy News as warranting investigation. The cause of the spike was not publicly explained by the USPTO or the defence agencies. Speculation ranges from expanded screening criteria to a surge of applications in sensitive technology areas -- quantum computing, advanced materials, directed energy -- but the official explanation is: there is no official explanation. Rescissions occur each year -- typically thirty to eighty -- but have not kept pace with new orders. The total has been climbing. The classified universe of suppressed inventions is expanding.
A fact buried in the FY2018 data deserves particular emphasis: of the approximately eighty-five new secrecy orders imposed that year whose government-interest status was recorded, forty-three affected inventions in which the government had no property interest. Forty-three orders imposed on private inventors. Not government contractors. Not researchers working under federal grants. Private citizens, filing patents on their own inventions with their own money, whose work was intercepted, classified, and suppressed. The majority of new secrecy orders in that fiscal year targeted private individuals.
Six thousand five hundred and forty-three inventions that the American public is not permitted to know about. The subjects of those orders are themselves secret. The inventors are gagged. The patents are withheld. The technologies do not reach the market. Whether any of those 6,543 orders covers an energy technology, a propulsion technology, or a gravity-related device is a question that the system is designed to make unanswerable.
There have been remarkably few legal challenges to secrecy orders. The fact is itself telling. The system's opacity and the criminal penalties for violation create powerful deterrents against litigation. An inventor under a secrecy order cannot publicise their situation without risking prosecution. They cannot organise with other affected inventors because they do not know who those inventors are. They cannot attract public support because they cannot describe what has been suppressed. The architecture ensures that each suppressed inventor is isolated, silenced, and alone. An inventor who avoids the patent system entirely -- who declines to file a patent and instead attempts to publish or commercialise independently -- may escape the Invention Secrecy Act. But an inventor who wants patent protection, which is essential for commercial development and investment, must enter the system. And for inventions with national security implications, the government can learn of them through other channels -- export control filings, research grant applications, defence contractor relationships, academic publications -- and intervene through other mechanisms described later in this chapter.
II. The CIA Act and the Precedent of Financial Opacity
Before FASAB Statement 56 legalised secret accounting for all agencies in 2018, one agency had been operating without financial accountability since 1949.
The Central Intelligence Agency Act of 1949, Public Law 81-110, signed on 20 June 1949, granted the CIA extraordinary operational authorities that no other civilian agency possessed. Section 10, codified at 50 U.S.C. 403j, authorised the Director of Central Intelligence to spend funds "without regard to the provisions of law and regulations relating to the expenditure of Government funds." The Director's voucher was sufficient authorisation. No standard procurement. No auditing. No accounting requirements. The CIA could spend any amount, on any programme, with no financial trace accessible to Congress, the Government Accountability Office, or the public. The DCI's personal voucher -- a single signature -- replaced the entire apparatus of federal financial oversight.
Section 5 exempted the CIA from disclosing the organisation, functions, names, official titles, salaries, or numbers of its personnel. Section 6 authorised the CIA to admit up to one hundred aliens per year to the United States outside normal immigration procedures, in the interest of national security. Section 8 established that CIA funds could be transferred to and received from other government agencies without regard to limitations on those agencies' budgets. The CIA's budget was hidden within other agencies' appropriations -- primarily the Department of Defence -- and the Director's personal voucher sufficed as the accounting record.
The CIA Act of 1949 is relevant to this chapter not because it is new -- it has been in force for over seventy-five years -- but because it established the precedent. It proved that the federal government could operate a major agency with no public financial accountability, for decades, without the system collapsing. The precedent was noted, absorbed, and eventually extended to the entire federal government through FASAB Statement 56.
The CIA's operational capabilities compound the financial opacity. MKUltra -- the CIA's programme of experiments on human subjects, authorised by Director Allen Dulles on 13 April 1953 -- demonstrates five facts about the agency's institutional capacity. First, the CIA funded secret research through eighty or more unwitting academic institutions for twenty years, operating 149 separate sub-projects across universities, hospitals, prisons, and pharmaceutical companies. The institutions included Harvard, Stanford, Columbia, MIT, UC Berkeley, and McGill University in Canada. Second, researchers often did not know who was really funding their work -- the front organisation mechanism, using intermediaries such as the Society for the Investigation of Human Ecology and the Geschickter Fund for Medical Research, meant that academics whose grants came from the CIA believed they were funded by private foundations. Third, the CIA was willing to destroy evidence: Director Richard Helms ordered destruction of the entire MKUltra archive in 1973. Only a cache of approximately twenty thousand financial documents survived because they had been misfiled in the financial records section -- they were accounting records rather than operational files, and the destruction order did not reach them. Those surviving records were discovered in 1977 during a FOIA request by journalist John Marks, who used them as the basis for The Search for the Manchurian Candidate (1979). Fourth, the programme operated for twenty years without public knowledge. Fifth, even after exposure through the Church Committee hearings in 1975 and the Senate Select Committee on Intelligence hearings in 1977, accountability was minimal -- no CIA officer was prosecuted. Helms received a suspended fine and a suspended sentence for lying to Congress about the programme's scope.
MKUltra is relevant not because of its specific subject matter but because it proves the mechanism. If the CIA could fund secret research through unwitting academic institutions for two decades using front organisations, if it could shape the intellectual discourse of the Western world for seventeen years through the Congress for Cultural Freedom -- a CIA-funded organisation active in thirty-five countries, publishing intellectual magazines including Encounter in London, Preuves in Paris, and Der Monat in Berlin, funding academic fellowships and cultural exchanges, all through intermediary foundations whose CIA backing was not exposed until 1967 -- if it could maintain over four hundred journalist relationships through Operation Mockingbird, as documented by Carl Bernstein in Rolling Stone in October 1977, with journalists at the New York Times, CBS, Time, and the Washington Post carrying out assignments for the Agency, the apparatus that CIA operative Frank Wisner called "the Mighty Wurlitzer" -- if it could operate an entire reconnaissance satellite programme (CORONA) in total secrecy from 1958 to 1995 -- then the institutional capability to fund, direct, and conceal research into any subject, including ether-adjacent physics, is documented fact, not speculation.
The CIA's Directorate of Science and Technology, formally established in 1963 by Director John McCone, consolidating the Office of Scientific Intelligence, the Office of Research and Development, and the technical collection systems, provides the institutional mechanism. The DS&T developed the U-2 reconnaissance aircraft under CIA contract with Lockheed's Skunk Works. It managed the A-12 OXCART programme -- the CIA's Mach 3+ reconnaissance aircraft, first flight 26 April 1962, one of the most tightly compartmented programmes in intelligence history. It operated the CORONA, GAMBIT, and HEXAGON satellite reconnaissance systems through the National Reconnaissance Office, a joint CIA-Air Force organisation whose very existence was classified from its creation in 1961 until 18 September 1992. The DS&T does not operate under peer review. Its assessments are classified. It can investigate any technology it chooses without public scrutiny or academic gatekeeping. If ether-based or vacuum-energy technologies showed any promise in classified testing, the DS&T would be the natural institutional home for assessing their potential.
The CIA's venture capital arm compounds the capability. In-Q-Tel, a nonprofit venture capital firm created by the CIA in 1999 under Director George Tenet, identifies and invests in early-stage technology companies whose innovations are relevant to intelligence needs. Its investment areas include artificial intelligence, quantum computing, advanced materials, and advanced sensors. Notable investments include Keyhole (whose technology became Google Earth) and Palantir Technologies. In-Q-Tel gives the CIA a direct mechanism for identifying, investing in, and potentially controlling emerging technologies. If a startup developed a breakthrough in vacuum energy or anomalous propulsion, In-Q-Tel could invest, gain board representation, and steer the technology into classified channels -- using the CIA Act of 1949's financial authorities to ensure the transaction left no public trace.
The CIA Information Act of 1984, Public Law 98-477, completed the opacity. This Act exempted CIA operational files from the search and review requirements of the Freedom of Information Act. The exempted categories are specific and comprehensive: files in the Directorate of Operations, the Directorate of Science and Technology, and the Office of Security. These are not marginal categories. They encompass the CIA's covert action files, its technology development files, and its security files -- the three categories most likely to contain any research into exotic technology. A FOIA requester cannot learn whether relevant records exist, let alone obtain them. The CIA can respond with a Glomar response -- neither confirming nor denying the existence of records. Any CIA research into exotic technology, vacuum energy, or ether-adjacent physics conducted through the Directorate of Science and Technology would be in files permanently invisible to the public.
III. FASAB Statement 56: Legalising the Secret Books
In October 2018, the Federal Accounting Standards Advisory Board -- the body that sets generally accepted accounting principles for the federal government, operating under the auspices of the Office of Management and Budget, the Government Accountability Office, and the Treasury Department -- issued Statement of Federal Financial Accounting Standards 56, titled "Classified Activities." It permits federal entities to modify their financial statements for reasons of national security.
"Modify" is the operative euphemism. The standard's text specifies two mechanisms. The first is component entity modifications: individual agencies modify their own financial statements to prevent disclosure of classified information. The second is government-wide adjustments: the consolidating entities, principally the Treasury, make adjustments at the government-wide level to reconcile modified component statements while still concealing the classified activities. The combined effect is a system where inaccuracy cascades from the agency level to the national accounts.
What the modifications permit, specifically: agencies may present numbers that are deliberately inaccurate in their public-facing financial reports. Expenses, assets, and liabilities can be reassigned to different budget categories. If one agency's classified programme spent five billion dollars, that five billion might appear in another agency's unclassified accounts as something else entirely. The amounts reported in individual line items can be altered -- not merely moved between categories but changed in their stated totals. The modifications need only be disclosed to the head of the entity and to a designated senior official. No public disclosure is required. No disclosure to Congress beyond existing classified reporting channels.
Before FASAB 56, agencies that modified financial statements to hide classified activities were technically violating federal accounting standards. This was not merely theoretical. The Department of Defence had failed to produce auditable financial statements since the Chief Financial Officers Act of 1990 first required them, and the impossibility of reconciling classified and unclassified accounts was cited as one reason.
After FASAB 56, they can comply with generally accepted accounting principles while presenting modified numbers. Audits can return "clean" findings even though the public numbers are intentionally inaccurate. Inter-agency transfers used to fund classified programmes can be hidden in plain sight. If the Department of Energy transfers two billion dollars to the Department of Defence for a classified programme, both agencies can adjust their books so that neither the source transfer nor the receiving entry appears. Congressional oversight is theoretically maintained through classified channels, but members of Congress working from unclassified budget documents are seeing false numbers and have no way to know it.
The drafting process deserves scrutiny. FASAB began deliberating on how to handle classified activities in financial statements in 2016-2017. The board noted that existing standards did not adequately address the tension between transparency requirements and national security classification. An exposure draft was issued for public comment. Transparency advocates argued that the standard would legalise what had previously been unauthorised manipulation of financial data, undermining the constitutional power of the purse. Some commenters noted that existing law -- classified annexes to appropriations -- already handled the national security and transparency tension and that SFFAS 56 went further than necessary. Government agencies, by contrast, generally supported the standard, arguing it was necessary to comply with audit requirements while protecting classified information. The vote on SFFAS 56 was not unanimous. At least one board member raised concerns or dissented, expressing worry that the standard could undermine public accountability and the integrity of government financial reporting. The dissent was noted in the record. The standard was adopted regardless.
The timing is what the Moynihan Commission would have called "suggestive."
In 2017, Dr Mark Skidmore, a professor of economics at Michigan State University specialising in public finance, working with Catherine Austin Fitts, a former Assistant Secretary of Housing and Urban Development under the George H.W. Bush administration and a former managing director at Dillon, Read & Co., published findings that twenty-one trillion dollars in undocumented adjustments appeared in the combined DOD and HUD financial statements between 1998 and 2015. These were accounting entries -- journal voucher adjustments -- that lacked adequate supporting documentation. The figure was cumulative over the period and dwarfed total federal expenditures, which run to roughly four to six trillion dollars per year. Skidmore has publicly stated that he was initially sceptical of Fitts's claims but was compelled by the government's own published data. His methodology was straightforward: he examined the agencies' own Office of Inspector General reports, which were publicly available on government websites. He used the government's own numbers. The twenty-one trillion figure was not an extrapolation or an estimate. It was the sum of undocumented adjustments that the OIG reports themselves identified.
In October 2018, FASAB 56 was issued.
Immediately following its issuance, the OIG reports that Skidmore had relied upon were removed from public access online. Some were subsequently restored, but the initial removal was documented.
Whether FASAB 56 was a direct response to the Skidmore findings or a pre-existing initiative that arrived at a convenient time is debated. The effect is not debated. FASAB 56 retroactively legalised the kind of financial statement modifications that the undocumented adjustments represented, and prospectively made it far more difficult for researchers to identify future discrepancies. The evidentiary window was closed. Fitts has described this as "deliberately closing the evidentiary window." Whether the metaphor is apt or not, the chronological sequence is: undocumented adjustments exposed (2017), the standard permitting such adjustments issued (2018), the source documents removed (2018). The sequence speaks for itself.
The Department of Defence has failed every comprehensive financial audit since the first was attempted in 2018, conducted by over 1,200 auditors from independent firms led by Ernst & Young. It failed in 2019. It failed in 2020, 2021, 2022, 2023, and 2024. Seven consecutive failures. It is the only major federal agency that has never passed a clean audit. DOD total assets are approximately 3.8 trillion dollars. The Pentagon has acknowledged it cannot fully account for how trillions of dollars have been spent. Some accounting experts argue that the undocumented adjustments reflect systemic accounting failures, duplicate entries, and the DOD's antiquated financial systems rather than hidden spending. The DOD's own response has been that its accounting systems are deeply flawed rather than that funds are hidden. Both interpretations acknowledge the same underlying fact: the Department of Defence cannot account for its money. Whether the cause is incompetence or concealment, the effect is identical: the money trail is broken.
The CIA Act of 1949 proved the concept: one agency could operate without financial accountability for decades. FASAB 56 extended the principle to the entire federal government. The intelligence community's acknowledged budget is itself substantial -- the National Intelligence Programme was disclosed at $52.6 billion for FY2013 through the Snowden revelations; the combined NIP and Military Intelligence Programme reached approximately $95 billion for FY2023. Fitts argues the actual classified spending is far larger than the acknowledged figures. The CIA Act of 1949 was the pilot programme. FASAB 56 was the rollout.
IV. The Classification Scale
The executive order governing national security classification is Executive Order 13526, signed by President Obama on 29 December 2009. It replaced EO 12958 (Clinton, 1995, as amended by Bush in EO 13292 in 2003). It is codified at 75 Federal Register 707 and available at the National Archives.
It establishes three levels -- Top Secret, Secret, and Confidential -- defined by the expected severity of damage from unauthorised disclosure: "exceptionally grave damage," "serious damage," and "damage" to national security, respectively. The standard is prospective and speculative: the classifier need not demonstrate actual damage, only a reasonable expectation of potential damage. The "could reasonably be expected to cause" language is a prediction, not a finding of fact -- an assessment of what might happen, not what has happened. The classifier's judgement is nearly unreviewable.
Section 1.4 defines what may be classified. The categories are: (a) military plans, weapons systems, or operations; (b) foreign government information; (c) intelligence activities, sources, or methods; (d) foreign relations or foreign activities; (e) scientific, technological, or economic matters relating to the national security; (f) programmes for safeguarding nuclear materials or facilities; (g) vulnerabilities or capabilities of systems, installations, infrastructure, projects, plans, or protection services relating to national security; and (h) the development, production, or use of weapons of mass destruction.
Category (e) is the broadest: "scientific, technological, or economic matters relating to the national security." This is an extraordinarily expansive category that could encompass virtually any advanced physics or engineering research deemed to have security implications. Gravity modification, vacuum energy extraction, exotic propulsion -- all would fall comfortably within category (e). The category does not require a weapon. It does not require a military application. It requires only a connection to "national security" -- a term the executive order does not define with precision.
Section 1.7 lists prohibitions: information may not be classified to conceal violations of law, prevent embarrassment, restrain competition, or prevent or delay the release of information that does not require protection. These prohibitions exist on paper. They have no effective enforcement mechanism. No classifier has been prosecuted for over-classification. No agency has been sanctioned. The prohibitions are aspirational language in a system that lacks the institutional will to enforce them.
Approximately 2,000 to 2,300 officials hold Original Classification Authority -- the power to make an initial determination that information warrants classification. Of these, a small number, roughly twenty to thirty, have Top Secret OCA delegated directly from the President. Several hundred have Secret OCA. The remainder have Confidential OCA. But this number is misleading, because the system's true scale operates through derivative classification.
Derivative classification is the process by which anyone with appropriate security clearance who handles existing classified information incorporates, restates, paraphrases, or generates new material based on it and applies the same classification markings. This does not require Original Classification Authority. It requires only clearance and access. Approximately four to five million people hold security clearances in the United States, including roughly 1.3 million with Top Secret clearances, according to recent Office of the Director of National Intelligence reports. The Information Security Oversight Office has reported that derivative classification decisions number in the tens of millions annually. In peak years through the mid-2000s to mid-2010s, the reported number exceeded fifty million per year. Even after reporting methodology changes reduced the count, the number remained in the tens of millions.
This is the amplification mechanism. A small number of original classification decisions propagate into millions of derivative classification actions, creating an ever-expanding body of classified material. If a single original classifier determines that a particular propulsion technology is classified under category (e), every subsequent document, report, email, memorandum, and technical assessment that references that technology is derivatively classified. The domain itself becomes classified. New work by cleared personnel in that domain is automatically classified. The classification propagates like a chain reaction. The number of originally classified documents -- approximately fifty to eighty thousand per year based on ISOO data -- is modest. The derivative classification machine amplifies that modest number into tens of millions of classification actions. No reliable figure exists for the total volume of classified documents in existence, but estimates range from hundreds of millions to over a billion pages across all agencies.
The system produces classified material far faster than it declassifies material. The National Declassification Centre processes forty to one hundred million pages per year for review, but many are deferred or exempted. Actual pages declassified and released are far fewer. The backlog of documents awaiting review was estimated at over four hundred million pages as of the 2010s. The 25-year automatic declassification provision in EO 13526, which requires review of records over twenty-five years old, has been repeatedly deferred through exemptions. The NDC has been chronically underfunded. The classified universe is expanding faster than it can be opened.
The Moynihan Commission -- the Commission on Protecting and Reducing Government Secrecy, chaired by Senator Daniel Patrick Moynihan -- reported in 1997 that "the classification system is used too much, at too high a level, for too long." Much classified information did not warrant protection. The system was "beset by inertia" -- it was easier to classify than to justify not classifying. "Secrets are not the enemy," the Commission concluded, "but their excessive accumulation is." The Commission recommended specific reforms: an "interest balancing" test weighing the public interest in disclosure against the security interest in secrecy, mandatory declassification timelines, and institutional mechanisms to enforce accountability. Most of its recommendations were not implemented. The system continued as before.
The 9/11 Commission found in 2004 that over-classification contributed to the intelligence failures preceding September 11 by impeding information sharing. "Current security requirements nurture over-classification and excessive compartmentation of information among agencies," the Commission reported. Classification barriers prevented analysts from connecting intelligence across agencies -- the very failure that allowed the attacks to succeed. The Commission's recommendations for reducing over-classification were partially implemented. The culture did not change.
Former ISOO director J. William Leonard publicly stated that "half or more" of classified material does not warrant classification. The Brennan Centre for Justice and other organisations have documented persistent over-classification. Former classification officials have testified that fifty to ninety per cent of classified information is over-classified. The people who run the system say it classifies too much. The system classifies more anyway. No one is punished for over-classifying. Careers are destroyed for under-classifying. The incentive structure points in one direction.
The born classified doctrine, established by the Atomic Energy Act of 1946, extends the principle to its logical extreme. Restricted Data -- all data concerning the design, manufacture, or utilisation of atomic weapons, the production of special nuclear material, or the use of special nuclear material in energy production -- is classified by its nature, automatically, from the moment of its existence. No government official needs to review the information and stamp it classified. A private scientist working in a private laboratory with no government funding who independently derives information about nuclear weapons design has generated Restricted Data and is legally prohibited from communicating it.
The Progressive case -- United States v. The Progressive, Inc., 467 F. Supp. 990 (W.D. Wis. 1979) -- illustrates what this means in practice. The magazine The Progressive prepared an article by Howard Morland titled "The H-Bomb Secret" based entirely on publicly available sources. The government obtained a prior restraint against publication, arguing the article contained Restricted Data. Judge Robert Warren granted the injunction -- the first prior restraint on publication upheld by a federal court since the Pentagon Papers case went the other way. The case was mooted when the same information was published elsewhere, but the precedent of judicial willingness to restrain publication of "born classified" material was established. The born classified system proved almost impossible to manage consistently. As historian Alex Wellerstein documented in Restricted Data: The History of Nuclear Secrecy in the United States (University of Chicago Press, 2021), the same information might be classified in one Department of Energy document and unclassified in another. The system was an improvisation during the Manhattan Project that became codified in law without full consideration of its long-term implications. Wellerstein documents how the classification system was used not just for security but for industrial policy -- controlling which companies and nations had access to nuclear technology. The scale of the nuclear classified archive alone is staggering: a 1973 review at Los Alamos National Laboratory alone examined 388,092 documents, declassified 234,315, and left over 150,000 classified. That was one laboratory, one review, when the programme was thirty years old. The total across the entire nuclear weapons complex over eighty years is almost certainly in the tens of millions of documents. Much has never been reviewed.
Formally, the born classified doctrine applies only to nuclear-related information. There is no equivalent statutory provision for exotic energy or advanced propulsion. But there are functional equivalents. The Invention Secrecy Act intercepts inventions at the patent office and imposes secrecy post hoc. Derivative classification under executive orders means that once a domain is classified, any new work in that domain by cleared personnel is derivatively classified, creating a propagating classification that functions similarly to born classified within the cleared community. The 2024 UAP Disclosure Act, as initially proposed by Senators Schumer and Rounds, contained language about "non-human intelligence" technologies that would have created new categories with born-classified characteristics, suggesting congressional awareness that such extension has occurred or could occur. The combined effect is a de facto born classified system for any technology domain that the defence establishment considers sensitive.
V. The SAP Framework: Programmes Within Programmes
Special Access Programmes are classification within classification -- an additional layer of restriction on top of the standard Top Secret/Secret/Confidential system. DOD Directive 5205.07, originally issued in 1995 and updated most recently in 2020, establishes the policy and procedures. A SAP is a programme established for a specific class of classified information that imposes safeguarding and access requirements exceeding those normally required for information at the same classification level. SAPs come in three functional categories under the directive: acquisition SAPs, which protect sensitive technology and programme information; intelligence SAPs, which protect intelligence activities; and operations and support SAPs, which protect especially sensitive military operations.
Access to SAP information requires more than a security clearance. It requires a specific need-to-know for the particular SAP, a special access approval separate from and in addition to the clearance, execution of a programme-specific non-disclosure agreement, and often a polygraph examination. The Secretary of Defence or designee establishes each SAP, and each has its own programme security officer and government programme manager. The compartmentation is designed to ensure that even within the cleared community, access to SAP information is restricted to the minimum number of individuals necessary.
An Acknowledged SAP is one whose existence is known but whose details are classified. The F-117 Nighthawk stealth fighter was an acknowledged SAP after its public unveiling in 1988: the public knew it existed, but its stealth technology remained SAP-classified. The total number of active acknowledged SAPs is not publicly reported in aggregate but is believed to number in the hundreds across the Department of Defence.
An Unacknowledged SAP is one whose very existence is classified. All aspects -- existence, purpose, budget, technology, personnel -- are classified. A cover story is maintained. If asked about the programme, authorised personnel may deny its existence. The number of active USAPs is itself classified.
Congressional notification is required under 10 U.S.C. Section 119, which mandates that the Secretary of Defence notify specified congressional defence committees of each SAP. The notification goes to the chairs and ranking members of the Senate Armed Services Committee, the House Armed Services Committee, the Senate Appropriations Committee (Defence subcommittee), and the House Appropriations Committee (Defence subcommittee) -- the "Gang of Eight" in the defence context.
A Waived SAP is the most restricted category. The Secretary of Defence may waive the congressional reporting requirement of 10 U.S.C. Section 119. The mechanism is as follows: the Secretary makes a determination that the normal notification would pose an "exceptionally grave" risk to national security. The notification to the full defence committees is waived. Instead, only a minimal notification is provided -- typically to the chairs and ranking members of the defence committees, four individuals, or potentially to no one outside the executive branch, depending on the interpretation of the waiver authority. The Secretary must notify Congress that a waiver has been invoked but need not identify the programme being waived. Congress is told that a waiver exists. It is not told what the waiver covers.
A waived SAP combined with FASAB 56 creates a programme where the existence is classified, congressional notification is waived, and the budget is buried in other appropriations under modified financial statements that are deliberately inaccurate. Only a handful of executive branch officials may know the programme exists. Article I, Section 9 of the Constitution requires that "a regular Statement and Account of the Receipts and Expenditures of all public Money shall be published from time to time." Waived SAPs, combined with FASAB 56, create a framework where this constitutional requirement is effectively nullified. The power of the purse -- the foundational check that the Constitution places on executive spending -- is suspended for any programme placed within this framework.
The declassified examples prove the system works at scale. The National Reconnaissance Office was created in 1961 as an unacknowledged programme. Its very existence was classified for thirty-one years, until 1992. Its annual budget during that period was in the billions -- at its peak in the 1960s, the NRO's budget exceeded NASA's. It built and operated the nation's spy satellites. For three decades, this entire agency officially did not exist. The B-2 Spirit stealth bomber was developed as a SAP with a total programme cost exceeding forty-four billion dollars in 1990s dollars -- demonstrating that SAPs can involve enormous expenditures hidden from public view for years. The F-117 Nighthawk achieved operational capability and flew combat missions years before the public was told it existed. These are not hypothetical capabilities of the system. They are its documented products.
The UAP Disclosure Act of 2023, initially proposed by Senators Schumer and Rounds as an amendment to the fiscal year 2024 National Defence Authorisation Act, is significant for what its language implies about what Congress believes exists within the SAP framework. The proposed legislation defined "technologies of unknown origin" and "non-human intelligence" as categories requiring disclosure. It defined "prosaic" versus "non-prosaic" attribution for aerospace technologies -- acknowledging the existence of technologies that cannot be attributed to known human capability. It created a review board modelled on the JFK Assassination Records Review Board to compel declassification, and included eminent domain provisions -- suggesting Congress believed that private aerospace contractors might hold exotic technologies and that government authority would be needed to compel their surrender. It targeted "legacy programmes" -- language implying long-running classified programmes that have operated for decades within the waived SAP framework. It required the executive branch to certify completeness of disclosure, with criminal penalties for false certification.
Senate Majority Leader Schumer stated on the Senate floor in July 2023: "The American public has a right to learn about technologies of unknown origin, non-human intelligence, and unexplained anomalous phenomena."
The Act was significantly weakened during the conference process, the House stripping the eminent domain and mandatory declassification provisions. Senator Schumer publicly attributed the opposition to defence industry interests. The iron triangle -- the self-reinforcing relationship between congressional committees that authorise defence spending, executive agencies that administer defence programmes, and defence contractors whose profits depend on continued government spending, the structure Eisenhower warned against in his 1961 farewell address -- protected itself. But the legislative language, drafted with access to classified briefings available to senators, constitutes an extraordinary statement about what senior legislators believe exists within the classified ecosystem. Senators do not draft eminent domain provisions for technologies they do not believe exist.
VI. NASA: The Shopfront
The National Aeronautics and Space Administration was created in the same eighteen-month window that consumed the gravitics programmes. As Chapter 7 documented, DARPA was established in February 1958 and NASA began operations on 1 October 1958, absorbing the National Advisory Committee for Aeronautics with its eight thousand employees and $117 million annual budget. The conventional explanation is that both were responses to Sputnik. The structural effect was something else.
The timeline of that eighteen-month period reveals how comprehensively the American aerospace landscape was restructured. Sputnik launched on 4 October 1957. ARPA was established by DOD Directive 5105.15 on 7 February 1958, signed by Secretary of Defence Neil McElroy. The National Aeronautics and Space Act was signed on 29 July 1958 as Public Law 85-568. NASA began operations on 1 October 1958. JPL was transferred from the Army to NASA on 3 December 1958. The Army Ballistic Missile Agency under Wernher von Braun began its transfer to NASA in 1959, formally completing on 1 July 1960 when von Braun's team became the nucleus of the Marshall Space Flight Centre. The legal infrastructure was already in place: the Invention Secrecy Act had been law since 1952, Executive Order 10290 had established the classification system for civilian agencies in 1951, DOD Directive 5200.1 had established classification authority for military-funded research in 1955, and the Atomic Energy Act had established the born classified regime in 1946. The public gravitics research programmes went silent across ten or more aerospace companies in this same window.
Section 102 of the National Aeronautics and Space Act established the foundational division: civilian space activities under NASA, military space activities under the Department of Defence. The Act declared that "activities in space should be devoted to peaceful purposes for the benefit of all mankind" -- but immediately qualified this by specifying that "activities peculiar to or primarily associated with the development of weapons systems, military operations, or the defense of the United States" were the Department of Defence's responsibility. The line between civilian and military was defined by whether an activity was "peculiar to or primarily associated with" defence -- and during the Cold War, this was an expansive category. The military's interpretation generally prevailed for anything touching national security.
Section 305 established the classification hierarchy. NASA could classify its own information, but it was subordinate to DOD on any information touching defence. NASA was required to defer to DOD classification decisions. NASA was required to make information available to DOD and other agencies on request. A Civilian-Military Liaison Committee was established to coordinate, but in practice the military's priorities dominated. The shopfront was open. The back room was locked. The same contractors -- Lockheed, Boeing, General Dynamics, North American -- worked both sides. JPL, NASA's premier planetary exploration laboratory, conducts classified work for the DOD and intelligence agencies and maintains personnel with NRO and other intelligence community clearances. Marshall Space Flight Centre in Huntsville sits adjacent to Redstone Arsenal, one of the Army's primary weapons development facilities. The overlap is structural, not incidental.
Eisenhower, who created this architecture, understood its implications. His farewell address of 17 January 1961 is remembered for the military-industrial complex warning. The less-quoted passage is more pointed: "The prospect of domination of the nation's scholars by Federal employment, project allocations, and the power of money is ever present and is gravely to be regarded. Yet, in holding scientific research and discovery in respect, as we should, we must also be alert to the equal and opposite danger that public policy could itself become the captive of a scientific-technological elite." The man who created both NASA and the civilian-military space division warned that the system he had built could become a mechanism for unaccountable power.
Within this structure, NASA's engagement with exotic propulsion has been a study in institutional minimalism.
The Breakthrough Propulsion Physics programme, run by Marc Millis at NASA's Glenn Research Centre in Cleveland, Ohio, was the only programme within the civilian aerospace agency that investigated whether the fundamental assumptions of space propulsion could be transcended. Millis, an aerospace engineer who joined NASA Glenn in 1982 with a physics degree from Georgia Tech, had spent years observing that while science fiction routinely depicted faster-than-light travel and gravity control, NASA had no programme investigating whether any physical basis existed for such technologies. In 1996, he successfully proposed an internal programme to investigate "breakthrough propulsion physics" -- defined as propulsion requiring no propellant mass, approaching or exceeding the speed of light, or tapping new energy sources.
It ran from 1996 to 2002. Its total lifetime budget was $1.6 million -- some sources cite $1.554 million -- approximately $228,000 per year. Against a NASA budget of approximately $14 billion per year, this represents roughly 0.0016 per cent -- 1.6 thousandths of one per cent. The only programme investigating breakthrough propulsion was funded at a level that would not cover a single engineer's fully burdened salary at a major aerospace contractor.
For comparison: the Mars Pathfinder mission cost $265 million. The Mars Exploration Rovers (Spirit and Opportunity) cost $820 million. The Cassini-Huygens mission cost $3.27 billion. The James Webb Space Telescope cost $10 billion. The Space Shuttle programme cost three to five billion dollars per year. The International Space Station cost over $100 billion, with the United States contribution at approximately $60 billion. The DIA's AAWSAP programme investigating anomalous aerospace phenomena received $22 million -- approximately fourteen times BPP's entire lifetime budget. NASA spent one thousand times more sending a probe to Saturn than it spent asking whether there was a better way to get there.
BPP investigated vacuum energy and zero-point energy extraction, including analysis of the Casimir effect and review of Puthoff's zero-point field theories. It investigated the Podkletnov gravity shielding claims: NASA Marshall Space Flight Centre attempted replication of Podkletnov's claimed gravitational shielding effect using rotating superconducting YBCO discs, with SCI Engineered Materials receiving $650,000 to manufacture the superconducting disc. The replication effort was plagued by technical difficulties -- the ceramic discs kept cracking -- and achieved only partial experimental conditions before funding ran out. The result was inconclusive: the experiment never achieved the conditions Podkletnov specified, so it neither confirmed nor refuted his claims. BPP funded small-scale testing of James Woodward's Mach effect thruster at California State University, Fullerton, which produced some anomalous thrust signals difficult to distinguish from systematic errors. It conducted theoretical analysis of the Alcubierre warp drive metric and assessed its energy requirements.
Its findings, documented in Millis and Eric Davis's Frontiers of Propulsion Science (AIAA, 2009), a peer-reviewed 739-page compendium published by the most prestigious professional organisation in aerospace engineering -- an organisation whose membership includes NASA's own engineers, which publishes the Journal of Spacecraft and Rockets and the Journal of Propulsion and Power, and which sets standards for aerospace engineering -- concluded that several research threads remained viable and warranted further investigation, that the research needed was at the intersection of general relativity, quantum mechanics, and thermodynamics, that the level of funding was inadequate to reach definitive conclusions on any major question, and that the field suffered from a lack of sustained, systematic investigation. BPP found that the questions were legitimate, the answers were inconclusive, and the programme was terminated before it could resolve anything.
The termination came in 2002, during the tenure of NASA Administrator Sean O'Keefe, a former Secretary of the Navy and OMB deputy director appointed by President George W. Bush. O'Keefe's mandate included imposing fiscal discipline after International Space Station cost overruns. BPP was not singled out by name in any public announcement. It was terminated as part of broader reductions in advanced concepts research at Glenn Research Centre -- the Glenn advanced concepts budget was cut to zero. Millis has stated publicly that there was no scientific review of BPP before its termination. It was a budget line that was eliminated. No successor programme was created. The institutional message was clear: NASA does not investigate breakthrough propulsion.
Millis left NASA and founded the Tau Zero Foundation in 2010 -- a private nonprofit dedicated to breakthrough propulsion research, funded by donations. The Tau Zero Foundation published research papers, hosted workshops, funded small research grants, maintained the BPP bibliography and research database, and advocated for renewed government investment. The world's foremost expert on breakthrough propulsion physics had to leave the government and create a private nonprofit to continue his work. The institutional message was explicit: NASA does not investigate whether the fundamental assumptions of space propulsion are correct.
Harold "Sonny" White's Eagleworks laboratory at Johnson Space Centre repeated the pattern at smaller scale. White, who held a PhD in physics from Rice University and served as Advanced Propulsion Theme Lead for the NASA Engineering Directorate, operated what amounted to an internal skunk works with minimal funding and staffing. His primary theoretical contribution was modifying the Alcubierre warp metric: Alcubierre's original 1994 paper had shown that general relativity permits a "warp bubble" solution, but the energy required was calculated at the mass-energy of Jupiter. White showed that by changing the geometry of the warp bubble to a thicker, more rounded configuration, the energy requirements could be reduced by many orders of magnitude -- potentially to the mass-energy equivalent of the Voyager 1 spacecraft. The Eagleworks laboratory did not have a dedicated NASA programme budget line; it operated within the Engineering Directorate's resources. The work was tolerated rather than championed by NASA leadership.
When White's team tested the EmDrive -- Roger Shawyer's truncated cone cavity in which microwaves resonate, which claimed to produce thrust without expelling propellant -- and published results in the AIAA Journal of Propulsion and Power in November 2016 reporting thrust of 1.2 plus or minus 0.1 millinewtons per kilowatt in vacuum conditions, the paper was peer-reviewed through AIAA's standard process and generated significant media attention. NASA maintained cautious distance, acknowledging the research was being conducted but not endorsing the results. Martin Tajmar's group at TU Dresden later found the apparent thrust attributable to thermal effects; when thermal effects were properly controlled for, the thrust signal disappeared.
When White published a paper in the European Physical Journal C in 2021 reporting that a custom Casimir cavity geometry produced an energy density distribution matching the requirements of the Alcubierre warp metric at nanoscale -- a finding that, if confirmed, would represent the first experimental identification of a real-world energy density distribution matching a warp metric -- NASA did not increase his funding. White left NASA for the Limitless Space Institute, a private nonprofit in Houston, continuing research into warp drive physics, advanced propulsion, and Casimir cavity experiments outside NASA's institutional framework. Another researcher pushed out of the government to continue his work with private money.
The pattern is now documented twice. Millis: NASA's exotic propulsion programme defunded; founds Tau Zero Foundation, a private nonprofit. White: NASA's exotic propulsion researcher leaves; joins Limitless Space Institute, a private nonprofit. In both cases, the researchers concluded that NASA was not an institution where exotic propulsion research could be sustained. In both cases, the path forward was outside the government. In both cases, the research continued at a fraction of the resources NASA could have provided.
NASA's 2023 UAP study -- responding to the most significant anomalous aerospace phenomenon in decades -- appointed one person: Mark McInerney, as NASA's Director of UAP Research. One person. NASA Administrator Bill Nelson had publicly acknowledged the existence of unexplained aerial phenomena, yet the agency's institutional response was a single appointment. The agency responsible for aerospace is not leading the investigation of anomalous aerospace phenomena. It is barely participating.
The privatisation of spaceflight has compounded the opacity. Under the Apollo model, NASA designed, built, and operated vehicles. Every dollar was subject to congressional appropriation and oversight. Technical designs were government property. FOIA applied to all programme documents. All technology developed was government-owned. Under the commercial model -- initiated under the Commercial Orbital Transportation Services programme in 2006 using Space Act Agreements rather than traditional contracts, with SpaceX receiving $396 million and Orbital Sciences receiving $288 million -- companies design and build vehicles with proprietary technology. NASA buys a service, not a vehicle. Companies retain intellectual property rights. Their internal research and development, processes, and proprietary information are not subject to FOIA. Companies can and do conduct classified DOD work using the same facilities, personnel, and technology base. Congressional oversight covers the NASA contract but not the company's broader operations. The shift from the Saturn V era to the SpaceX era changed more than the cost of launch. It changed who can ask questions about what happens in orbit.
The X-37B -- a Boeing-built autonomous spaceplane operated by the Space Force -- has spent over nine hundred days in orbit on classified missions whose objectives are undisclosed. It has flown multiple missions, each classified, each of extended duration, each invisible to the public. Lockheed Martin builds NASA's Orion crew vehicle and also builds classified military aerospace systems through its Skunk Works division. The same company, one side public, the other side dark. When public capability is moved behind corporate walls, public accountability goes with it. The FOIA request that would have reached a NASA-designed vehicle cannot reach a SpaceX-designed vehicle. The privatisation did not merely change who builds the hardware. It changed who can ask questions about it.
NASA defines what "serious" aerospace research looks like. Universities align their research programmes with what NASA funds. Doctoral students work on problems NASA considers legitimate. Peer review panels are populated by researchers formed within this system. Proposals for exotic physics research are evaluated by reviewers who consider it fringe. The few programmes that investigate exotic physics are funded at negligible levels and eventually defunded. The defunding is cited as evidence that the ideas have no merit. The cycle repeats. The system is self-reinforcing regardless of whether anyone intends it to be.
VII. The Ning Li Case
The architecture described in the preceding sections is abstract until it is applied to a person. Ning Li's story is the application.
Li received her doctorate in physics from UC Santa Barbara in the mid-to-late 1980s. She joined the University of Alabama in Huntsville around 1989 as a research scientist. UAH had -- and has -- a significant aerospace and physics research presence tied to the nearby Marshall Space Flight Centre and Redstone Arsenal, making it a natural environment for gravity-related research. With Douglas Torr, a senior professor in the UAH physics department, she published a series of papers that constitute the core of the Li-Torr theoretical programme:
In Physical Review D, volume 43, number 2, published 15 January 1991: "Effects of a Gravitomagnetic Field on Pure Superconductors." This was the foundational paper. Li and Torr argued that the coupling between lattice ions and Cooper pairs in a superconductor would amplify gravitomagnetic effects enormously compared to normal matter. The theoretical basis drew on the DeWitt minimal coupling prescription -- Bryce DeWitt had shown in the 1960s that the Schrodinger equation in a gravitational field acquires terms analogous to electromagnetic vector potentials. Li and Torr extended this to the BCS theory of superconductivity, arguing that the macroscopic quantum state of a superconductor amplified the gravitomagnetic coupling by a factor related to the ratio of the ion mass to the Cooper pair mass and the density of the condensate.
In Physical Review B, volume 46, number 9, published 1 September 1992: "Gravitational effects on the magnetic attenuation of superconductors." This extended the theoretical framework, examining how gravitational fields interact with the magnetic properties of superconducting materials.
In Foundations of Physics Letters, volume 6, number 4, August 1993: "Gravitoelectric-electric coupling via superconductivity." This proposed that time-dependent gravitomagnetic fields could generate detectable gravitoelectric fields, and that superconductors provided a mechanism for coupling these to measurable electromagnetic signals.
In standard general relativity, gravitomagnetism is an extraordinarily weak effect. The frame-dragging measured by Gravity Probe B required extraordinary precision to detect -- roughly 37 milliarcseconds per year of precession. Gravitomagnetic effects in laboratory-scale systems are suppressed by factors of order v/c and GM/rc-squared, making them approximately ten to the power of negative twenty to negative thirty times electromagnetic forces. Li and Torr's predictions implied enhancement factors that, if correct, would bring gravitomagnetic effects into the measurable regime. This is what made the work both extraordinary and controversial: it predicted effects many orders of magnitude larger than standard physics allowed.
In 1999, Edward Harris of the University of Tennessee published a critique in Foundations of Physics Letters, volume 12, number 2, pages 201-208: "Comments on 'Gravitoelectric-Electric Coupling via Superconductivity' by Douglas Torr and Ning Li." Harris re-examined the Li-Torr derivation and identified what he considered fundamental errors in their treatment of the gravitomagnetic coupling. His central claim was that Li and Torr had made an error in the application of the minimal coupling prescription, effectively double-counting certain contributions and arriving at a coupling strength enormously larger than a correct calculation would yield. Harris contended that the "gravitomagnetic" London moment predicted by Li and Torr arose from confusing canonical momentum with kinetic momentum in the gravitational context. In his analysis, when the calculation was done correctly, the enormous enhancement factor disappeared, and the gravitomagnetic coupling reverted to the standard GR value -- approximately twenty to thirty orders of magnitude smaller than predicted. This is not a minor quantitative disagreement. It is the difference between a measurable laboratory effect and one that is completely undetectable with any foreseeable technology.
Li appears to have maintained that her theoretical framework was correct and that Harris had made incorrect assumptions about the nature of the Cooper pair coupling. No definitive published rebuttal in a peer-reviewed journal has been identified in open sources. By 1999, Li was already transitioning away from UAH and toward her private company.
The Department of Defence did not agree with the mainstream perspective. In 2001 -- two years after Harris's critique -- the Army Research Office awarded Li's private company, AC Gravity LLC, incorporated in the state of Alabama with a Huntsville address, a grant of $448,970 to investigate AC gravity effects using rotating superconducting materials. The grant came after the critique was published. The DOD funded the work knowing that the mainstream had rejected the theoretical foundation. The stated objective was to develop and test an experimental apparatus based on Li's predictions about generating gravitomagnetic fields using rapidly spinning superconducting discs. Standard DOD research grant deliverables would include periodic progress reports, a final technical report, and publication of results subject to classification review. Either DOD reviewers disagreed with Harris, or the DOD had independent reasons -- possibly experimental results from other programmes, possibly intelligence about foreign research -- for continuing the investment.
Li received a Top Secret security clearance, approximately 2001-2002. The clearance required a Single Scope Background Investigation, which typically takes six to twelve months and involves interviews with associates, financial review, and depending on the specific access level, a polygraph examination. A physics professor at a state university receiving TS clearance for this particular line of research implies that the DOD considered the work to have national security implications. This is consistent with two interpretations: the DOD believed the technology could lead to breakthrough propulsion or weapons-relevant capabilities, or the DOD wanted to ensure that if the research produced positive results, those results could be immediately classified. Both interpretations lead to the same structural outcome: the work was positioned to disappear into the classified domain.
In May 2003, a communication attributed to Li -- reported secondhand through Tim Ventura of American Antigravity, a website and media project that tracked exotic propulsion research in the early 2000s, and never independently verified in its original form -- referenced "eleven kilowatts of output effect" from her experimental apparatus. In context, this would mean that her device was producing a gravitational force effect equivalent to eleven kilowatts of energy output -- an extraordinary claim. If the claim was genuine, it would represent a macroscopic gravitational effect produced by a tabletop device -- one of the most significant physics results of the century. Its secondhand nature and the absence of any corroborating publication mean it cannot be taken as established fact. It exists in the record as a data point: an unverified claim of extraordinary results, followed by total silence.
Then silence. Li ceased publishing. She ceased attending conferences. She stopped responding to inquiries from journalists and fellow researchers. This silence began around 2003 and continued for the rest of her life. Her company, AC Gravity LLC, became inactive. Alabama business records show no further contracts, publications, or public activity after approximately 2003-2004.
Multiple individuals, including Tim Ventura and other researchers in the exotic propulsion community, filed FOIA requests with the Army Research Office and other DOD entities for the grant's deliverables and final report. The results were uniformly negative -- either "no records found" or denial on classification grounds. The grant results were never published. The standard deliverables of a DOD research contract -- progress reports, a final technical report, publication of results -- simply vanished. The specific FOIA language -- "no records found" -- may mean the records do not exist, or that they are held in SAP systems that the FOIA office does not search. The distinction is invisible to the requester.
The parallel to Tesla, documented in Chapter 6, operates with unsettling precision. Both were researchers whose work centred on the interaction between electromagnetism and gravity. Both attracted government interest. Both entered the classified domain. Both produced claims of significant results. Both were silenced. Both had their work disappear from the public record. Tesla died alone in Room 3327 of the Hotel New Yorker. Li was struck by a vehicle near the UAH campus in 2014. Her husband died in 2015. She was diagnosed with Alzheimer's disease. She died on 27 July 2021, in Huntsville, Alabama, having never spoken publicly about her research again.
Her son, speaking to the Huntsville Business Journal in a 2023 investigative article titled "Solving the Mystery of Huntsville's Brilliant Scientist Disappearing," confirmed several facts. She had remained in Huntsville the entire time -- she never moved to China, never defected, never disappeared in the dramatic sense. This directly contradicted the claim made by Jack Sarfatti, a physicist in the exotic physics community, who circulated around 2008 the rumour that Li had "defected to China" or been recruited by the Chinese government. The rumour was false, debunked by her own son. She had experienced significant personal tragedy: her husband's death, the vehicle accident, her own declining health and Alzheimer's diagnosis. Her son confirmed the biographical details while being unable or unwilling to discuss the classified work. No one the journal could reach had information about the outcome of the DOD-funded research. The grant results remain unaccounted for in the public record.
A published physicist who produced peer-reviewed work in the most respected physics journals, who attracted nearly half a million dollars in government funding, who received the highest security clearance, who reportedly claimed significant experimental results -- and whose work was absorbed into the classified domain so completely that not a single deliverable has ever surfaced. The most parsimonious explanation is classification absorption: the research produced results -- positive, negative, or ambiguous -- that the DOD classified, and Li, bound by her clearance and NDAs, could not discuss the work. Her personal circumstances compounded the silence. The grant deliverables exist in classified archives and have not been released. This is consistent with standard DOD research practices. Many defence-funded research programmes simply disappear into classified channels, especially if the results have weapons implications. The Ning Li case is notable not because the pattern is unprecedented, but because the subject matter -- gravity modification -- is so inherently extraordinary that the classification creates a much more dramatic narrative gap than it would for a new radar absorbing material.
The pattern: published, funded, cleared, results claimed, silence, FOIA denied, death.
VIII. The Invisible College
While Ning Li's case shows the architecture absorbing and silencing a single researcher, a parallel structure reveals something different: a network that has operated within the classified ecosystem for over fifty years, publishing selectively in peer-reviewed journals while conducting unknown classified work. Jacques Vallee borrowed Robert Boyle's seventeenth-century term for the informal network that became the Royal Society: the invisible college.
Harold "Hal" Puthoff received his PhD in electrical engineering from Stanford in 1967 and co-authored Fundamentals of Quantum Electronics (Wiley, 1969), a widely used textbook. Before Stanford, he worked at the National Security Agency as a civilian, in signals intelligence and communications related to his electronics and physics background. The exact nature of his NSA work remains classified. His NSA background gave him established relationships within the intelligence community and an active security clearance -- both of which proved critical for everything that followed.
In 1972, Puthoff joined SRI International -- formerly Stanford Research Institute, a major defence contractor and research institution in Menlo Park, California -- and, with Russell Targ, a laser physicist who had co-invented the FM laser, co-founded what became the Stargate programme -- the United States government's investigation of remote viewing, the claimed ability to perceive distant or hidden targets through means not explainable by known physics. The programme was funded by the CIA and later the Defence Intelligence Agency, operating under a succession of code names: SCANATE (1972-1973, CIA-funded), GONDOLA WISH and GRILL FLAME (1977-1983, DIA-funded), CENTER LANE (1983-1985), SUN STREAK (1985-1990), and STAR GATE (1990-1995). Puthoff directed the SRI component from 1972 to approximately 1985.
Twenty-three years of continuous funding for research that the mainstream scientific establishment considered impossible. Total funding exceeded twenty million dollars across the programme's lifetime. The programme was initiated after intelligence reports suggested the Soviet Union was spending heavily on psychic research -- the Soviet "psychotronics" programme. The most operationally significant claimed success: Pat Price's remote viewing of a Soviet nuclear test facility at Semipalatinsk, reportedly verified by satellite imagery. The programme produced thousands of sessions with varying degrees of apparent accuracy. Statistical analyses showed hit rates above chance for the best viewers in controlled conditions.
When the programme was declassified in 1995, the CIA commissioned an external review by the American Institutes for Research, with statistician Jessica Utts of UC Davis and sceptic psychologist Ray Hyman of the University of Oregon providing duelling assessments. Utts concluded that the statistical evidence for a real effect was overwhelming, with effect sizes comparable to many accepted medical interventions. Hyman, while acknowledging the statistical anomalies, argued that methodological issues prevented definitive conclusions. The CIA's administrative conclusion was that even if a real effect existed, it was not reliable or specific enough for operational intelligence use, and the programme was terminated. The intelligence community funded it for twenty-three years because operational results appeared to justify continued investment. Whatever the ultimate scientific verdict on remote viewing, the programme establishes a documented precedent: United States intelligence agencies were willing to fund, for over two decades, research that mainstream science rejected.
In 1985, Puthoff left SRI and founded EarthTech International, a for-profit research company in Austin, Texas, and the Institute for Advanced Studies at Austin, a nonprofit research institute operating in tandem with EarthTech. IASA served as the academic and nonprofit arm; EarthTech was the commercial and contracting arm. Together they constituted Puthoff's research enterprise, focused on advanced physics concepts including zero-point energy, advanced propulsion, and fundamental physics related to the quantum vacuum.
His published physics work is where the connection to the ether framework becomes explicit.
In Physical Review A, volume 39, number 5, published 1 March 1989: "Gravity as a zero-point-fluctuation force." Puthoff proposed that gravity could be understood not as curvature of spacetime but as a residual electromagnetic effect of the quantum vacuum zero-point field. The core idea: all matter interacts with the zero-point electromagnetic field; this interaction produces a mutual attraction between massive bodies that mimics Newtonian gravity. He derived Newton's law of gravitation from zero-point field interactions using stochastic electrodynamics formalism. The paper was controversial but was published in a top physics journal and remains widely cited.
In the same journal, volume 40, number 9, published 1 November 1989: "Source of vacuum electromagnetic zero-point energy." This companion paper addressed the self-consistency of the zero-point field by arguing it could be understood as a self-regenerating cosmological phenomenon -- the quantum vacuum maintains itself through a feedback mechanism involving interactions with matter.
In 1994, with Bernhard Haisch and Alfonso Rueda, he published in Physical Review A, volume 49, number 2: "Inertia as a zero-point-field Lorentz force." This landmark paper -- the "HRP" paper -- proposed that inertia is not an intrinsic property of matter but arises from the interaction of accelerating charged particles with the quantum vacuum zero-point field. When a charged particle accelerates through the zero-point field, it experiences a frequency-asymmetric radiation pressure that manifests as the inertial force F = ma. Newton's second law is derived from electrodynamics rather than being a fundamental axiom. If correct, mass is not intrinsic but emergent from vacuum interactions, and inertia could in principle be manipulated by modifying vacuum interactions. The paper was widely covered in the popular science press and generated significant debate. Y. Levin published a critique in 2009 arguing that the HRP calculation contained errors in the treatment of radiation reaction; Haisch and Rueda published responses defending their approach. The debate remains technically unresolved in the open literature.
Most significantly, in Foundations of Physics, volume 32, number 6, June 2002: "Polarizable-Vacuum (PV) Approach to General Relativity." Puthoff showed that all standard results of general relativity -- gravitational redshift, light bending, perihelion precession, Shapiro delay -- could be reproduced by treating the vacuum as a polarisable medium: essentially, a dielectric ether. In his model, the speed of light varies with gravitational potential because the vacuum itself has variable electromagnetic properties -- permittivity and permeability change near mass. The metric tensor of general relativity is replaced by a scalar index of refraction of the vacuum. The refractive index K varies as K = exp(2GM/rc-squared) near a mass M. Gravity is not spacetime curvature but a change in the electromagnetic properties of the vacuum medium. Puthoff was careful to present this as a "pedagogical tool" and "engineering model," but the implications are clear. The PV model is an ether theory published in a peer-reviewed physics journal by a researcher with decades of intelligence community connections and security clearances. Puthoff has used the phrase "the modern ether" and acknowledged that his PV model recovers the same physics as general relativity but with an ontology closer to the nineteenth-century ether tradition.
Eric Davis received his PhD in astrophysics from the University of Arizona in approximately 1991, focusing on general relativity, quantum field theory in curved spacetime, and exotic spacetime metrics. He was a long-time research physicist at Puthoff's Institute for Advanced Studies, focusing on traversable wormholes, vacuum energy extraction, and exotic spacetime metrics. He authored the Air Force Research Laboratory study "Teleportation Physics Study" (AFRL-PR-ED-TR-2003-0034, 2004), which surveyed teleportation concepts including quantum teleportation, wormhole-based transit, and -- controversially -- psychic teleportation. He authored several of the thirty-eight Defence Intelligence Reference Documents commissioned under AAWSAP, including those on traversable wormholes and stargates, antigravity for aerospace applications (surveying Podkletnov, Li-Torr, and the Tajmar experiments), and concepts for extracting energy from the quantum vacuum. These DIRDs were produced as unclassified reference documents for DIA personnel and were eventually released through FOIA.
The Wilson-Davis memo requires separate attention. This document -- also called the Wilson Notes -- allegedly records a meeting on 16 October 2002 between Vice Admiral Thomas R. Wilson, who had served as Director of the Defence Intelligence Agency (1999-2002) and prior to that as J-2 (Director of Intelligence) for the Joint Chiefs of Staff, and Eric Davis. According to the document, Wilson described how, in 1997, while serving as J-2, he attempted to gain access to a Special Access Programme involved in the analysis of recovered technology "not of this Earth." He claimed he identified the programme but was denied access despite his senior rank. He was told by programme managers that the programme existed outside normal congressional oversight. He was threatened with career consequences if he pursued the matter. The document was found among the papers of Dr Edgar Mitchell, Apollo 14 astronaut, after Mitchell's death in 2016. Wilson has publicly denied the meeting took place. Davis has been circumspect, declining to discuss it in at least one interview, citing its sensitive nature, but has not issued a categorical denial. The document remains unverified. It has been discussed in congressional hearings on UAP. Several members of Congress have referenced alleged reverse-engineering programmes in ways consistent with its contents. It has not been authenticated by any official investigation, but neither has it been definitively debunked. Its relevance here is structural: if its contents are accurate, it describes a SAP operating outside both congressional oversight and the normal chain of command -- the waived SAP framework described in Section V of this chapter, in operation.
Davis subsequently moved to the Aerospace Corporation, a federally funded research and development centre primarily serving the Space Force and the National Reconnaissance Office, approximately 2017-2019. His specific work there is not publicly detailed, consistent with the classified nature of much of the organisation's work.
Christopher "Kit" Green, MD, PhD, served in the CIA's Office of Scientific Intelligence as chief of the Biomedical Sciences Division, a career spanning the 1970s through the 1990s, evaluating SRI remote viewing results from the intelligence consumer side. He held TS/SCI clearance. After leaving the CIA, he worked at General Motors in their research and development division in medical diagnostics and biomedical engineering, before returning to government-adjacent work. He later authored DIRD #10 on "Field Effects on Biological Tissues" for the AAWSAP programme, examining cases where individuals reported physical effects from encounters with anomalous phenomena -- burn injuries, radiation-like symptoms, and neurological effects reported by close-encounter witnesses.
Jacques Vallee, born 24 September 1939 in Pontoise, France, holds a PhD in computer science from Northwestern University (1967) and was a member of the team at SRI International that helped develop ARPANET -- the precursor to the internet. He is a genuine pioneer of computer networking, a fact often overlooked in discussions of his subsequent career as the most intellectually significant figure in UAP research. Vallee was at SRI concurrently with Puthoff in the 1970s. He was a consultant to the AAWSAP programme and has discussed his involvement in Forbidden Science, Volume 4 (2019). His parallel career as a venture capitalist in Silicon Valley gave him financial independence from academic or government funding, allowing him to pursue unconventional research without institutional pressure.
Robert Bigelow, a Las Vegas billionaire whose fortune came from the Budget Suites of America hotel chain, funded the National Institute for Discovery Science (1995-2004), where Puthoff, Davis, Green, and Vallee converged. NIDS was a privately funded scientific organisation dedicated to investigating anomalous phenomena, with particular focus on UAP, anomalous phenomena at Skinwalker Ranch (a 512-acre property in Uintah County, Utah, that Bigelow purchased in 1996), and survival of consciousness. NIDS deployed scientific teams to the ranch, documented in Hunt for the Skinwalker by Colm Kelleher and George Knapp (2005). Kelleher was NIDS's deputy administrator. Several NIDS personnel -- Kelleher, Green, Puthoff, Davis, Vallee -- subsequently became involved in Bigelow Aerospace Advanced Space Studies, creating direct institutional continuity.
Bigelow then won the DIA's AAWSAP contract through BAASS, valued at approximately $22 million, deploying the same network to execute it. BAASS commissioned the thirty-eight DIRDs, investigated UAP reports including deploying field teams, conducted investigations at Skinwalker Ranch, compiled databases, and supported the medical study of physiological effects on witnesses. In a 2017 interview with Lara Logan on CBS's 60 Minutes, Bigelow stated: "I'm absolutely convinced" that extraterrestrials exist and that UFOs have visited Earth. When asked if it was risky to say so, he replied: "I don't give a damn. I don't care. It's not going to make a difference. It's not going to change the reality of what I know."
The thirty-eight DIRDs commissioned under this programme constitute the United States government spending twenty-two million dollars on vacuum engineering, antigravity, wormholes, and exotic propulsion -- while publicly maintaining that these are fringe topics unworthy of serious investigation. The list includes studies on advanced space propulsion based on vacuum (spacetime metric) engineering, traversable wormholes, negative energy, antigravity for aerospace applications, quantum vacuum energy extraction, field effects on biological tissues, metallic spintronics, advanced nuclear propulsion, and many others. These are not the topics of a programme that considers the physics settled.
The institutional lineage is a documented through-line spanning five decades: SRI International (1972-1985) to EarthTech/IASA (1985-present) to NIDS (1995-2004) to BAASS/AAWSAP (2008-2012) to TTSA (2017-2021) to the Aerospace Corporation, the Sol Foundation, and current positions. In October 2017, To The Stars Academy of Arts and Sciences was founded by Tom DeLonge, with Puthoff as founding member and Vice President of Science and Technology, Luis Elizondo (who ran AATIP inside the Pentagon), Steve Justice (from Lockheed Skunk Works), and Chris Mellon (former Deputy Assistant Secretary of Defence for Intelligence). TTSA represented the "disclosure" phase -- the network partially surfacing into public view. Puthoff continues at EarthTech and advises the Sol Foundation, an academic organisation founded in 2023 to study UAP implications. The network persists.
The same core individuals. Multiple institutional shells. Multiple funding streams: CIA, DIA, Army, Air Force, NASA's BPP programme, and private capital. PhD-holding scientists with genuine credentials in physics, engineering, astrophysics, and medicine. Active security clearances maintained throughout, allowing them to move between classified and unclassified domains.
What they publish is speculative but within the bounds of theoretical physics: vacuum energy, polarisable vacuum models, stochastic electrodynamics, wormhole metrics. What they may know from classified work is unknown. If the Wilson-Davis memo is authentic, Davis at minimum was told of the existence of recovered non-human technology. If AAWSAP investigations produced significant results -- and the programme ran for years with twenty-two million dollars in funding, suggesting the sponsors saw value -- the network members have access to data that is not in the public domain.
The gap between the two -- between their public theoretical work on vacuum energy and polarisable vacua and whatever motivated their decades of intelligence-funded investigation -- is the gap the classification architecture creates. Their published work could be understood as the unclassified edge of a classified research programme: the part that can be published without revealing what motivated the theoretical investigation. The invisible college has operated for half a century at the intersection of legitimate physics, intelligence funding, and classification. Its public output treats the vacuum as an engineerable medium. Its classified output is, by design, invisible.
IX. Sealing Every Exit
The individual components of the architecture -- the Invention Secrecy Act, the CIA Act, FASAB 56, the classification scale, the SAP framework, NASA's structural subordination, ITAR and EAR export controls, FOIA exemptions -- are not a collection of independent provisions. They are an interlocking system. Each component reinforces the others. Each seals an exit that another leaves open.
An invention is made. A private citizen, working with private money, discovers something.
If the inventor files a patent: the Invention Secrecy Act intercepts it. The PSCRL flags the relevant technology category. A Type III General Secrecy Order suppresses all disclosure indefinitely. The inventor is silenced under criminal penalty. The technology can be absorbed into a classified programme. The international dimension ensures no foreign patent rights are preserved. The compensation mechanism is structurally inadequate. The inventor is alone, gagged, and without effective recourse.
If the inventor seeks government funding: the research is conducted under contract with classification provisions. The results are classified under Executive Order 13526, category (e): "scientific, technological, or economic matters relating to the national security." Derivative classification propagates the secrecy to every subsequent document in the domain. The researcher receives a security clearance and signs programme-specific NDAs. The deliverables vanish into classified archives. FOIA requests return "no records found."
If the inventor attempts to publish: ITAR's International Traffic in Arms Regulations control defence-related technical data. The United States Munitions List -- twenty-one categories covering defence articles, services, and related technical data -- includes launch vehicles, guided missiles, spacecraft, and military electronics. ITAR controls not just physical items but information: blueprints, drawings, plans, instructions, and any information required for the design, development, production, or modification of defence articles. If the research is deemed to relate to any of these categories, publication requires a State Department licence. The "deemed export" rule means that even sharing information with a foreign colleague in an American laboratory can constitute an illegal export. Penalties under the Arms Export Control Act reach one million dollars per violation and twenty years imprisonment. The Export Administration Regulations under the Commerce Department's Bureau of Industry and Security control dual-use technologies. The Commerce Control List includes advanced materials and superconductors, advanced sensors, quantum devices, and propulsion systems and components. The emerging and foundational technologies provision of the 2018 Export Control Reform Act specifically directs BIS to identify and control novel technologies before they become widely available. Between ITAR and EAR, every channel of international dissemination is covered. A researcher who avoids the patent system and thus the Invention Secrecy Act but attempts to publish internationally or collaborate with foreign scientists can be caught by export controls. The two systems cover different channels but reach the same result.
If anyone files a FOIA request: the Freedom of Information Act provides nine exemptions. Exemption (b)(1) covers classified national security information -- the most frequently invoked exemption for defence and intelligence records. Exemption (b)(3) covers information exempt from disclosure by other statutes -- and the National Security Act, the Atomic Energy Act, the Invention Secrecy Act, and the CIA Information Act all provide independent statutory bases for withholding. This exemption is particularly powerful because it does not require balancing the public interest against the withholding: if the statute applies, the records are exempt, period. Exemption (b)(5) protects pre-decisional deliberative communications and is frequently used as a catch-all; courts have repeatedly criticised its overuse, but agencies continue to invoke it broadly. Exemption (b)(4) protects trade secrets and commercial information submitted by private contractors -- relevant when defence contractors' proprietary data is sought. The Glomar response denies even the existence of records. Agencies can submit classified declarations to the court ex parte and in camera -- privately, without the requester seeing them. The requester has the burden of litigation. The agency has the resources of the federal government. The statutory response deadline is twenty business days, but agencies routinely take months or years, particularly for national security-related requests.
If Congress investigates: Waived SAP status limits notification to a handful of individuals -- or to none. FASAB 56 hides the spending. The executive branch can invoke executive privilege, classification, and sources-and-methods protections. The Senate's proposed eminent domain provisions in the UAP Disclosure Act -- designed to compel private contractors to surrender exotic technologies -- were stripped from the final legislation. Congress attempted to open a door. The door was closed in conference. The iron triangle -- congressional committees, executive agencies, defence contractors -- protected the architecture. Senator Schumer attributed the opposition publicly to defence industry interests. The mechanism by which the House stripped the provisions was the conference process, where the interests Eisenhower warned against in 1961 operate with maximum leverage and minimum public visibility.
If the courts are asked to intervene: the state secrets privilege, established in United States v. Reynolds (345 U.S. 1, 1953), allows the government to block discovery of classified information in litigation. The government asserts that disclosure would harm national security, and judges typically defer to executive branch classification determinations. The privilege has been invoked in cases ranging from surveillance to rendition to contract disputes involving classified programmes. In Reynolds itself, the Supreme Court accepted the government's claim that Air Force accident reports contained state secrets. Decades later, when the reports were declassified, they contained no secrets -- only evidence of government negligence that the state secrets privilege had been used to conceal. The privilege protects not just secrets but the government's interest in avoiding accountability. The courts are not a check on the system. They are part of it.
If the inventor avoids all channels -- does not file a patent, does not seek government funding, does not publish, does not collaborate internationally: the technology remains in a private laboratory, unknown to the patent office, unknown to ITAR, unknown to FOIA. But it also remains unknown to the world. An inventor who avoids every point of interception achieves secrecy -- self-imposed secrecy indistinguishable from the government-imposed variety. The architecture does not need to suppress what is never shared. Its genius is that it makes sharing impossible and not sharing pointless. The inventor's only paths to commercialisation, investment, and recognition -- patents, publications, grants, collaborations -- are the paths the architecture controls.
The programme operates indefinitely. Funded through modified appropriations. Staffed by cleared personnel bound by programme-specific non-disclosure agreements. Invisible to the public. Partially invisible to Congress. Protected at every point of potential disclosure.
The system does not require a conspiracy. It does not require a secret committee meeting in a darkened room, issuing orders to suppress ether physics. Each component operates independently under its own legal authority, administered by its own bureaucracy, justified by its own logic. The USPTO screens patents under the Invention Secrecy Act because it is required to. Defence agencies review flagged applications because they are authorised to. FASAB 56 permits modified financial statements because the board adopted the standard. Classification officers classify because the executive order directs them to. FOIA officers deny requests because the exemptions apply. Judges defer to classification because precedent instructs them to. Export control officers restrict dissemination because the regulations mandate it.
No coordination is required because the architecture is self-coordinating. Every component points in the same direction: toward concealment. An invention enters the system at any point -- a patent filing, a government grant, a journal submission, a FOIA request, a congressional inquiry, a court case -- and the system routes it toward classification and silence. The components interlock like the gears of a machine. The machine runs whether anyone is watching it or not.
X. The Weight of the Architecture
The cumulative weight of the architecture documented in this chapter can now be assessed.
A statute enacted in 1952 gives the government power to suppress any patent -- including purely private inventions -- under criminal penalty, indefinitely, with no meaningful judicial review. Over 6,500 orders are active. The subjects are secret. The inventors are gagged. In a single fiscal year, forty-three new orders were imposed on private inventors with no government connection. The FY2024 spike to 356 new orders -- nearly triple the historical average -- remains unexplained. The PSCRL's screening criteria capture not just weapons but energy technology, efficiency thresholds, and propulsion concepts. The international dimension strips not just domestic but global patent rights. The compensation mechanism is structurally designed to fail.
An accounting standard issued in 2018 legalises deliberately false government financial statements. The Department of Defence cannot account for trillions of dollars and has failed seven consecutive annual audits. Twenty-one trillion dollars in undocumented adjustments were identified in the government's own records. The OIG reports documenting those adjustments were removed from public websites. The predecessor, the CIA Act of 1949, proved the model could work for seventy-five years. The CIA's financial opacity is compounded by operational capabilities demonstrated across eighty institutions for twenty years (MKUltra), across thirty-five countries for seventeen years (Congress for Cultural Freedom), across over four hundred journalists for decades (Operation Mockingbird), and across an entire satellite reconnaissance programme for thirty-seven years (CORONA). The CIA Information Act of 1984 made the Directorate of Science and Technology's files permanently invisible to FOIA search.
A classification system produces tens of millions of derivative classification decisions annually, expanding faster than it declassifies, with half or more of its classified material over-classified according to the people who run the system. The Moynihan Commission recommended reforms. They were not implemented. The 9/11 Commission found over-classification contributed to intelligence failures. The culture did not change. The born classified doctrine automatically classifies an entire domain of physics. Functional equivalents extend the principle beyond nuclear.
A Special Access Programme framework creates programmes whose existence is classified, whose congressional notification is waived, and whose budgets are hidden in modified appropriations. Entire agencies have operated within this framework for thirty-one years without the public knowing they existed. Tens of billions of dollars have been spent in total secrecy. Senators with access to classified briefings drafted legislation implying the existence of exotic technologies held by private contractors within this framework, including eminent domain provisions to compel their surrender. Those provisions were stripped.
The civilian space agency is subordinate to the military on classification and has funded its only exotic propulsion programmes at negligible levels before defunding them entirely. Its two leading exotic propulsion researchers both left the government to found private nonprofits in order to continue their work. Its response to the UAP phenomenon has been a single appointment.
Export controls criminalise the international sharing of defence-related technical data. FOIA exemptions deny access to records through multiple overlapping statutory bases. The state secrets privilege blocks discovery in litigation. The Glomar response denies even the existence of records. The iron triangle protects the architecture from legislative reform.
A physicist at the University of Alabama published peer-reviewed papers on gravity modification, received a DOD grant and Top Secret clearance, reportedly claimed significant results, and vanished into complete silence. Not a single deliverable has surfaced. She died in obscurity. The FOIA requests returned nothing.
A network of PhD-holding scientists with intelligence community connections has operated for fifty years across multiple institutional shells, funded by the CIA, the DIA, the Army, the Air Force, and private capital, publishing ether-adjacent physics in peer-reviewed journals while conducting classified work that remains invisible. Their public output treats the vacuum as an engineerable medium. Their classified output is, by the system's design, unknowable.
Each component is a matter of public law. None is secret. The architecture is visible. Its applications are not.
This is the machine that makes the silence permanent. Chapter 6 documented how finance killed Tesla's technology. Chapter 7 documented how an entire industry was silenced in eighteen months. This chapter has documented the legal, institutional, and financial apparatus that ensures the silence persists — not for years but for decades, not for one inventor but for thousands, not through a single act of suppression but through an interlocking system that routes every form of disclosure toward classification and every form of inquiry toward denial.
The three preceding chapters have documented three dimensions of suppression operating in concert: financial (Morgan's withdrawal of capital), industrial (the simultaneous silencing of ten companies), and institutional (the legal architecture of permanent concealment). The ether was not merely removed from the textbooks. It was removed from the patent office, from the budget, from the FOIA response, from the courtroom, from the congressional record, and from the public's ability to know that any of this occurred.
The question that remains is not how the suppression was accomplished. The question is why. Who benefits from an architecture that suppresses energy technology, silences propulsion research, and classifies entire domains of physics? The answer lies not in the security apparatus but in the financial system it protects — a system built on energy scarcity, sustained by the petroleum economy, and defended by the most powerful financial institutions on Earth. That system is the subject of the chapters that follow.