September 11 · evidence ledger

Evidence

All 39 evidence records in this case's ledger. Every record carries an explicit direction, keeps what the source states separate from what we infer, and links its source and the claims it bears on. Supporting and undermining evidence get the same structure and the same seriousness.

Supports21 Undermines11 Qualifies3 Context4

Supports

21 records · strongest first

Records that favor a rung of the featured hypothesis.

Supports· decisiveSEP-E001

The federal report: seven hours of fire, total collapse, first known instance

what the source states: "On September 11, 2001, WTC 7 endured fires for almost seven hours, from the time of the collapse of the north WTC tower (WTC 1) at 10:28:22 a.m. until 5:20:52 p.m., when it collapsed. This was the first known instance of the total collapse of a tall building primarily due to fires."
what we infer: Establishes the event and its duration from the investigating agency's own final report. The "first known instance" framing was accurate when written and requires updating in light of the 2017 Tehran collapse (SEP-E018), which is a different point from being wrong.
limitations: The collapse time is video-derived; seismic records place it at 5:20:33 p.m., a measurement-chain difference of about twenty seconds.

S. Shyam-Sunder (2008)AI-verified citationSEP-C001

Supports· decisiveSEP-E003

The free-fall interval, stated in the federal report's own principal findings

what the source states: "A more detailed analysis of the descent of the north face found three stages: (1) a slow descent with acceleration less than that of gravity that corresponded to the buckling of the exterior columns at the lower floors, (2) a freefall descent over approximately eight stories at gravitational acceleration for approximately 2.25 s, and (3) a decreasing acceleration as the north face encountered resistance from the structure below." Elsewhere: "In Stage 2, the north face descended at gravitational acceleration, as the buckled columns provided negligible support to the upper portion of the north face."
what we infer: The free-fall interval is not a challenger's finding contested by the agency; it is the agency's own published measurement, added after public comment, and it is accompanied by the agency's explanation for it. Both facts belong in any honest account of this dispute.
limitations: A measurement of one visible face from video, not of the structure as a whole.

S. Shyam-Sunder (2008)AI-verified citationSEP-C002

Supports· decisiveSEP-E008

The court record: 68,500 files withheld, upheld under the statute

what the source states: "On July 9, 2009, the Director of NIST issued a formal finding pursuant to the NCSTA providing that the public disclosure of certain information gathered by NIST in connection with its investigation into the collapse of the WTC 7 building 'might jeopardize public safety.'" On the counts: "Overall, NIST produced a total of 25,644 data files responsive to Quick's request and withheld a total of 68,500 data files." On the rationale: "the withheld data files contain information that could be used to predict the collapse of a building and, if made available to a person with the appropriate level of expertise, would provide instruction to individuals wanting to learn how to simulate building collapses and how to most effectively destroy large buildings."
what we infer: Establishes that the central mechanism claim cannot be independently recomputed, and that this was a deliberate decision with a stated rationale a court accepted. The rationale is coherent; it is also exactly what makes SEP-C003 unauditable, and both things are true at once.
limitations: The opinion's own figures do not reconcile: 25,644 released plus 68,500 withheld against 91,114 described as potentially responsive.
The court decided the lawfulness of the withholding, not the adequacy of the underlying model.

U.S. District Court for the District of Columbia (2011)AI-verified citationSEP-C005

Supports· decisiveSEP-E035

The inspector general on edited air-quality assurances

what the source states: "However, when EPA made a September 18 announcement that the air was 'safe' to breathe, it did not have sufficient data and analyses to make such a blanket statement." And: "Furthermore, The White House Council on Environmental Quality influenced, through the collaboration process, the information that EPA communicated to the public through its early press releases when it convinced EPA to add reassuring statements and delete cautionary ones." In detail: "As a result of the White House CEQ's influence, guidance for cleaning indoor spaces and information about the potential health effects from WTC debris were not included in EPA's issued press releases."
what we infer: Documented official deception on a health-relevant matter, established by the agency's own inspector general — and therefore the calibration anchor for this whole case. It licenses withholding presumptive deference from official reassurance. It does not license inferring concealment about the attacks themselves, and the same report is evidence that the government's internal oversight worked.
limitations: The copy hosted on the agency's own site is a 32-page excerpt; the complete report is the repository version cited in the source record.

U.S. Environmental Protection Agency, Office of Inspector General (2003)AI-verified citationSEP-C024

Supports· strongSEP-E002

FEMA 2002: broken water main, no manual firefighting, and FEMA's own low confidence

what the source states: "WTC 7 collapsed approximately 7 hours after the collapse of WTC 1. Preliminary indications were that, due to lack of water, no manual firefighting actions were taken by FDNY." On the water supply: "It appeared that water on site was limited due to a 20-inch broken water main in Vesey Street." And on its own conclusion: "The specifics of the fires in WTC 7 and how they caused the building to collapse remain unknown at this time. Although the total diesel fuel on the premises contained massive potential energy, the best hypothesis has only a low probability of occurrence. Further research, investigation, and analyses are needed to resolve this issue."
what we infer: Independent of the later federal investigation and earlier than it, this corroborates the unfought-fire condition. Its final passage is also the honest baseline for how open the question was in 2002 — the first official study rated its own best explanation improbable and asked for more work.
limitations: Explicitly preliminary, conducted without the site access or resources of the later investigation.
Its diesel-fuel hypothesis was superseded by the later investigation, which found fuel-oil fires played no role.

Federal Emergency Management Agency (2002)AI-verified citationSEP-C001 SEP-C003

Supports· strongSEP-E004

The official collapse sequence: thermal expansion, a girder walking off its seat, Column 79 buckling

what the source states: "The initiating local failure that began the probable WTC 7 collapse sequence was the buckling of Column 79. This buckling arose from a process that occurred at temperatures at or below approximately 400 °C (750 °F), which are well below the temperatures considered in current practice for determining fire resistance ratings associated with significant loss of steel strength." And: "Fire-induced thermal expansion of the floor system surrounding Column 79 led to the collapse of Floor 13 ... the floor beams on the east side of the building expanded enough that they pushed the girder spanning between Columns 79 and 44 to the west on the 13th floor. ... This movement was enough for the girder to walk off of its support at Column 79."
what we infer: The mechanism is specific, falsifiable, and requires no melted steel — the report puts the highest column temperatures near 300 °C. Its specificity is what makes SEP-C005 consequential: a precisely stated mechanism that cannot be recomputed is a testable claim nobody can test.
limitations: Described as the probable sequence, not an observed one; it is the output of models whose data files were largely withheld.
The girder designation used throughout public argument does not appear in this report, which refers only to the girder spanning Columns 79 and 44.

S. Shyam-Sunder (2008)AI-verified citationSEP-C003

Supports· strongSEP-E009

The agency's own account of why no WTC 7 steel was preserved

what the source states: "Steel samples were removed from the site before the NIST investigation began. In the immediate aftermath of Sept. 11, 2001, debris was removed rapidly from the site to aid in recovery efforts and to facilitate emergency responders' efforts to work around the site. Once it was removed from the scene, the steel from WTC 7 could not be clearly identified. Unlike the pieces of steel from WTC 1 and WTC 2, which were painted red and contained distinguishing markings, WTC 7 steel did not contain such identifying characteristics."
what we infer: A candid agency acknowledgment of a permanent evidentiary loss. It forecloses the class of test — direct metallurgical examination of identified WTC 7 steel — that could have discriminated between the competing mechanisms, which is why no future disclosure can fully resolve this branch of the case.
limitations: The agency explaining a gap in its own evidentiary base.

National Institute of Standards and Technology (2021)AI-verified citationSEP-C006

Supports· strongSEP-E013

The fireproofing counterfactual, with the report's own hedge

what the source states: "The WTC towers likely would not have collapsed under the combined effects of aircraft impact damage and the extensive, multi-floor fires that were encountered on September 11 2001, if the thermal insulation had not been widely dislodged or had been only minimally dislodged by aircraft impact." And the companion finding: "In the absence of structural and insulation damage, a conventional fire substantially similar to or less intense than the fires encountered on September 11, 2001, likely would not have led to the collapse of a WTC tower."
what we infer: The mechanism is a conjunction — impact damage AND stripped insulation AND multi-floor fire — and the report says "likely" in both statements. Restatements on both sides routinely drop that hedge in opposite directions, one to claim certainty and the other to claim overreach.
limitations: An inference from tuned impact and fire models, not a measurement of the buildings as they stood.
Quoted from a PDF whose text layer carries OCR artifacts; the printed text is unambiguous.

S. Shyam-Sunder (2005)AI-verified citationSEP-C011

Supports· strongSEP-E014

The agency's own statement that collapse progression was not modeled

what the source states: "Once the collapse had begun, the propagation of the collapse was readily explained without the same complexity of modeling." The supporting argument given is a capacity comparison: "The vertical capacity of the connections supporting an intact floor below the level of collapse was adequate to carry the load of 11 additional floors if the load was applied gradually and 6 additional floors if the load was applied suddenly (as was the case)."
what we infer: Confirms directly that no end-to-end simulation of either tower's total progression was produced, and that the agency's position rests on an engineering judgment supported by a capacity argument. That is a defensible position and a real gap in the auditable record — which is why SEP-C012 grades the existence of a model rather than the physics.
limitations: An agency FAQ restating a position, not a technical publication.

National Institute of Standards and Technology (2021)AI-verified citationSEP-C012

Supports· strongSEP-E020

Peer-reviewed finding: unusually high put buying, consistent with informed trading

what the source states: "Examination of the option trading leading up to September 11 reveals that there was an unusually high level of put buying. This finding is consistent with informed investors having traded options in advance of the attacks."
what we infer: Independent peer-reviewed confirmation that the anomaly is real as a statistical matter. Note the published claim's exact form: consistency with informed trading, which is not a demonstration of it and cannot identify a trader.
limitations: Verified from the abstract; the paywalled full text was not obtained, so no methodological detail is quoted.
Statistical abnormality cannot identify a trader or a motive.

Allen M. Poteshman (2006)AI-verified citationSEP-C015

Supports· strongSEP-E023

The August 6 brief: patterns of suspicious activity, no time, place, or method

what the source states: The brief item, reproduced in full in the report, is headed "Bin Ladin Determined To Strike in US" and states: "Nevertheless, FBI information since that time indicates patterns of suspicious activity in this country consistent with preparations for hijackings or other types of attacks, including recent surveillance of federal buildings in New York." The report notes it was "the 36th PDB item briefed so far that year that related to Bin Ladin or al Qaeda, and the first devoted to the possibility of an attack in the United States."
what we infer: Simultaneously the strongest documented warning and a demonstration of how unspecific the warnings were: it names no date, target, or method. A correction this case had to make: the brief says preparations for "hijackings or other types of attacks," not for a hijacking, and the narrower misquotation makes the warning sound more precise than it was.
limitations: Redacted material is bracketed in the reproduction; the unredacted item has not been published.
The report's own body text and the reproduced document capitalize the title differently.

National Commission on Terrorist Attacks Upon the United States (2004)AI-verified citationSEP-C018 SEP-C019

Supports· strongSEP-E025

The inspector general on analytical failure: 'a significant intelligence failure'

what the source states: "What we do know is that the FBI was not adequately analyzing information for the purpose of drawing conclusions and making predictions. This was a significant intelligence failure, which hindered the chances of the FBI being able to detect and prevent the September 11 attacks." On the two future hijackers: "In sum, we found individual and systemic failings in the FBI's handling of information regarding the Hazmi and Mihdhar matter. The FBI had at least five opportunities to learn about their presence in the United States and to seek to find them before September 11, 2001." And on the flight-school detention: "we did not find that any employees committed intentional misconduct, or violated established FBI policies or practices, or attempted to deliberately sabotage the Moussaoui case. But the performance of several individuals involved with the case was lacking."
what we infer: Documents the failure mode at each node in detail, and specifically finds no intentional misconduct where the complicity hypotheses would require it. Five missed opportunities on the same two men is the signature of an institution not connecting information, not of one protecting a plot.
limitations: An agency's inspector general reviewing that agency.
The 'significant intelligence failure' judgment addresses analytical practice generally and is frequently narrowed in quotation to a single episode.

U.S. Department of Justice, Office of the Inspector General (2004)AI-verified citationSEP-C018 SEP-C019

Supports· strongSEP-E026

The Commission's staff on a scramble launched after a phantom aircraft

what the source states: "The report of American 11 heading south as the cause of the Langley scramble is reflected not just in taped conversations at NEADS, but in taped conversations in FAA centers, on chat logs compiled at NEADS, continental region headquarters, and NORAD, and in other records. Yet this response to a phantom aircraft, American 11, is not recounted in a single public timeline or statement issued by FAA or DOD. Instead, since 9/11, the scramble of the Langley fighters has been described as a response to the reported hijacking of American 77, or United 93, or some combination of the two." On notice: "NEADS received notice of the hijacking nine minutes before it impacted the North Tower. The nine minutes notice was the most the military would receive that morning of any of the four hijackings."
what we infer: Documented divergence between the recorded evidence and every public official account for nearly three years. This is negative evidence about official candor — and, in the same breath, the strongest evidence against the shootdown hypotheses, since fighters chasing an already-crashed aircraft were not intercepting anything.
limitations: A commission's staff characterizing agencies' public statements; the agencies' own explanations appear in the same hearing's testimony.
Establishes inaccuracy, not intent.

National Commission on Terrorist Attacks Upon the United States (2004)AI-verified citationSEP-C020

Supports· moderateSEP-E011

Witness accounts of explosion-like sounds at WTC 7

what the source states: A broadcast clip caption: "NY1 reporter Gigi Stone Woods on September 11, 2001: 'All of a sudden, a loud, incredibly loud explosion.'" The same page reports her later confirmation by email on January 4, 2021: "All I remember is that we were reporting near building 7 [sic] heard a loud explosion and people yelling to run and we all did." A police officer's account: "And the whole time you're hearing THOOM, THOOM, THOOM, THOOM, THOOM. So I think I know an explosion when I hear it, you know?"
what we infer: Establishes that loud explosion-like sounds were reported by people present. It establishes nothing about their cause: what a layperson identifies as an explosion is not a measurement, and collapsing structure, failing transformers, and falling debris all produce comparable reports.
limitations: Compiled and published by an advocacy organization; the accounts are collected in support of a conclusion.
The organization's characterization of what the investigating agency claimed about witness reports is its own and is not adopted here — see SEP-E012 for the agency's actual statement.

Architects & Engineers for 9/11 Truth (2021)AI-verified citationSEP-C008

Supports· moderateSEP-E018

Tehran 2017: fire-induced total collapse of a steel-framed high-rise

what the source states: "On January 19, 2017 an accidental fire, which started by an electric shorting, resulted in the complete collapse of the 16-storey Plasco Building in Tehran. Twenty-two people including sixteen firefighters were killed in that incident." The investigation summarizes "the sequence of events from the start of the fire to final collapse, almost 3.5 h later, and the inferred scenario explaining how the structure collapsed completely."
what we infer: Establishes the mechanism class — a steel-framed high-rise can suffer total fire-induced collapse — which the proponent literature's central prior denied. It does not establish structural similarity to WTC 7, and this case does not extend it that far. Its significance is chronological: the 2015 and 2018 proponent arguments were built on a pre-2017 record and require updating, which is different from being refuted.
limitations: A different structural system, fire load, and damage history; the comparison is at the level of mechanism class only.
Verified from the published abstract; the full text was not obtained.

M. T. Ahmadi (2020)AI-verified citationSEP-C003 SEP-C001

Supports· moderateSEP-E019

The exchange volumes: put-to-call ratios of 12:1 and 6:1

what the source states: "On 6 and 7 September 2001, the Chicago Board Options Exchange handled 4,744 put options for United Airlines' stock, translating into 474,000 shares, compared with just 396 call options, or 39,600 shares. On a day that the put-to-call ratio would normally have been expected to be roughly 1:1 (no negative news stories about United had broken), it was instead 12:1." And: "On 10 September 2001, another uneventful news day, American Airlines' option volume was 4,516 puts and 748 calls, a ratio of 6:1 on yet another day when by rights these options should have been trading even."
what we infer: The anomaly at the observation rung, in the specific numbers that circulate. Its interpretation is a separate question graded at SEP-C016.
limitations: A secondary carrier: these figures descend from September 2001 press reporting of exchange data, and no primary exchange record was obtained — so this evidence and that press coverage are one source, not two.

David Mikkelson (2001)AI-verified citationSEP-C015

Supports· moderateSEP-E021

A second peer-reviewed method finds abnormal pre-attack trades across airlines and banks

what the source states: "Comparable abnormal trades have been found for American Airlines, United Airlines and Boeing (and to a lesser extent for Delta Air Lines and KLM) before the terrorist attacks of 9/11, and are discussed in the separate appendix."
what we infer: A second independent method reaching a compatible conclusion, which is why the anomaly is graded well supported rather than merely reported. Recorded with a sourcing precision this case had to correct: the granular figures widely attributed to this paper — a specific count of trades and specific dollar gains — are not in its published body but in its working-paper version and separate appendix, and are not quoted here.
limitations: The published paper treats the September 11 trades only summarily; its detailed findings sit in a separate appendix this case has not verified.
Detects statistical abnormality, not identity or intent.

Marc Chesney (2015)AI-verified citationSEP-C015

Supports· moderateSEP-E027

The recordings behind the corrected chronology

what the source states: Reporting built on roughly thirty hours of air-defense operations-floor recordings obtained from the Pentagon, the article documents that "the fact that the fighters have been launched in response to the phantom American 11—rather than American 77 or United 93—is referred to six more times on Nasypany's channel alone," and that by the time the sector learned of a bomb report on United 93, "everyone on board is already dead."
what we infer: Independent corroboration from the primary recordings that the phantom pursuit was pervasive on the operations floor rather than an isolated misunderstanding, and that the aircraft had already crashed before the sector had actionable information about it.
limitations: Journalism: the recordings are the primary source and this article is the route to them.
The summary phrase often attributed to this piece — that the military never had a hijacked airliner in its sights — does not appear in it and is not quoted as the author's words.

Michael Bronner (2006)AI-verified citationSEP-C020

Supports· moderateSEP-E029

Five witnesses barred from testifying, and a committee chairman's objection

what the source states: "I was surprised to find that the Department of Defense has ordered five key witnesses not to testify, some of them military, some civilian, all working for the Department of Defense. That looks to me as if it may be obstruction of the Committee's activities, which is something we will have to determine."
what we infer: The department prevented sworn public testimony on the question a year before its own inspector general resolved it internally. That is a real transparency failure and it is not evidence about what the program found — an important distinction this case keeps, because the barring is often offered as though it were.
limitations: The chairman's characterization of possible obstruction is his assertion at the time, not a finding.

U.S. Senate Committee on the Judiciary (2005)AI-verified citationSEP-C021

Supports· moderateSEP-E030

Finding 20 of the withheld section, with the inquiry's own caveat

what the source states: "20. Finding: While in the United States, some of the September 11 hijackers were in contact with, and received support or assistance from, individuals who may be connected to the Saudi Government." The same section states: "it should be clear that this Joint Inquiry has made no final determinations as to the reliability or sufficiency of the information regarding these issues that we found contained in FBI and CIA documents. It was not the task of this Joint Inquiry to conduct the kind of extensive investigation that would be required to determine the true significance of any such alleged connections to the Saudi Government."
what we infer: The two sentences must travel together, and in public argument they almost never do. The finding reports contact and assistance from individuals who "may be" connected; the caveat states the inquiry did not assess whether the underlying information was reliable. Quoting either alone misrepresents the document.
limitations: Redacted at publication to protect sources and methods, so the released text is incomplete by construction.
Reports raw investigative leads from agency files rather than adjudicated findings.
The document is an image-only scan; these quotes were read by optical character recognition and warrant a human re-read before promotion.

U.S. House Permanent Select Committee on Intelligence (2016)AI-verified citationSEP-C022 SEP-C023

Supports· weakSEP-E039

The original nanothermite paper: red/gray chips with a low-temperature exotherm

what the source states: "We have discovered distinctive red/gray chips in all the samples we have studied of the dust produced by the destruction of the World Trade Center." "The red material contains grains approximately 100 nm across which are largely iron oxide, while aluminum is contained in tiny plate-like structures. Separation of components using methyl ethyl ketone demonstrated that elemental aluminum is present." "When ignited in a DSC device the chips exhibit large but narrow exotherms occurring at approximately 430 ˚C, far below the normal ignition temperature for conventional thermite." "The red portion of these chips is found to be an unreacted thermitic material and highly energetic."
what we infer: The affirmative case for the claim, recorded with the same structure as the evidence against it. Its central assertion — that elemental aluminum is present — is precisely what the independent analysis at SEP-E017 failed to find, so the two findings are in direct factual conflict rather than talking past each other. Graded weak on venue and provenance grounds rather than on the chemistry, which no third laboratory has adjudicated.
limitations: The journal's editor-in-chief resigned over its publication, stating it was printed without her permission and should never have appeared in that journal — while also noting the subject lay outside her expertise, so her objection went to process rather than chemistry.
Samples were collected privately by correspondents of the authors, with no documented chain of custody from the site.
Never independently reproduced on custody-controlled material.
Verified from the published abstract; the full text was not obtained, so no methodological detail is quoted.

Niels H. Harrit (2009)AI-verified citationSEP-C014

Undermines

11 records · strongest first

Records that count against a rung — shown with the same structure and seriousness.

Undermines· strongSEP-E012

The acoustic threshold: what a column-severing charge would have sounded like

what the source states: "Blast events did not play a role in the collapse of WTC 7. ... Blast from the smallest charge capable of failing a critical column (i.e., Column 79) would have resulted in a sound level of 130 dB to 140 dB at a distance of at least half a mile if unobstructed by surrounding buildings (such as along Greenwich Street or West Broadway). This sound level is consistent with standing next to a jet plane engine and more than 10 times louder than being in front of the speakers at a rock concert. There were no witness reports of such a loud noise, nor was such a noise heard on the audio tracks of video recordings of the WTC 7 collapse."
what we infer: A falsifiable published prediction that the recorded evidence does not meet. Read precisely, the agency denies blast-LEVEL noise, not booms — its public FAQ separately answers a responder who reported two loud booms by arguing the reported levels do not match a column-failing blast. Both the prediction and its exact scope matter, and public argument on both sides misstates them.
limitations: The threshold is the investigation's own modeling, not an independent measurement.
It assumes an unobstructed path in an environment full of obstructions, which the report acknowledges by naming specific sightlines.

S. Shyam-Sunder (2008)AI-verified citationSEP-C009 SEP-C008

Undermines· strongSEP-E015

No pre-collapse seismic signal, and no corroborating evidence of demolition

what the source states: "There were no seismic signals that occurred prior to the initiation of the collapse of either tower. The seismic record contains no evidence that would indicate explosions occurring prior to the collapse of the towers."
what we infer: Directly contradicts the explosive-demolition account, which requires detonations immediately before initiation. Recorded here with its correct attribution: this negative finding is the agency's, citing the seismic record — it is commonly and wrongly attributed to the observatory that produced the underlying measurements.
limitations: The agency's characterization of the seismic record rather than the seismologists' own publication, which this case has not obtained.

National Institute of Standards and Technology (2021)AI-verified citationSEP-C013

Undermines· strongSEP-E017

The dust chips: no elemental aluminum — and not the building's primer either

what the source states: "The red/gray chips found in the WTC dust at four sites in New York City are consistent with a carbon steel coated with an epoxy resin that contains primarily iron oxide and kaolin clay pigments." And: "There is no evidence of individual elemental aluminum particles of any size in the red/gray chips, therefore the red layer of the red/gray chips is not thermite or nano thermite." On the primer question: "The composition of the red/gray chips found in this study (epoxy resin with iron oxide and kaolin pigments) does not match the formula for the primer paint used on iron column members in the World Trade Center towers" and "none of the 177 different coatings are a match for the red layer coating found in this study."
what we infer: Elemental aluminum is not optional for the thermite claim, so its absence in an independent analysis of custody-archived dust is a direct contradiction. But the same report refutes the popular counter-claim: it explicitly did NOT identify the chips as the towers' column primer. What the chips actually are remains unidentified in the published literature — a real loose end that both camps paper over.
limitations: Never published in a peer-reviewed journal — the same venue objection made against the paper it rebuts.
Prepared for a private client and surviving only on third-party mirrors, one of them adversarial to the hypothesis tested.
Analyzed the analyst's own archived samples, not the samples used in the paper it rebuts, so it is an independent test rather than a re-examination.

James R. Millette (2012)AI-verified citationSEP-C014

Undermines· strongSEP-E022

The Commission's trace-out: 95 percent of the airline puts to one investor with no conceivable ties

what the source states: "Some unusual trading did in fact occur, but each such trade proved to have an innocuous explanation. ... A single U.S.-based institutional investor with no conceivable ties to al Qaeda purchased 95 percent of the UAL puts on September 6 as part of a trading strategy that also included buying 115,000 shares of American on September 10. Similarly, much of the seemingly suspicious trading in American on September 10 was traced to a specific U.S.-based options trading newsletter, faxed to its subscribers on Sunday, September 9, which recommended these trades. ... These investigators have found that the apparently suspicious consistently proved innocuous."
what we infer: Specific, plausible, and unverifiable from outside. The footnote names neither the investor nor the newsletter nor the underlying agency files, and its cited authorities are interviews and an internal memorandum that remain non-public — so the trace-outs cannot be audited. That is why SEP-C016 is graded weakly supported rather than contradicted: the explanation is probably right and nobody outside the government can check it.
limitations: The investigating bodies' own summary of their own unpublished work product.
The footnote asserts that each unusual trade proved innocuous while publishing the reasoning for only two examples.

National Commission on Terrorist Attacks Upon the United States (2004)AI-verified citationSEP-C016

Undermines· strongSEP-E024

The joint inquiry's first factual finding, in both of its halves

what the source states: "1. Finding: While the Intelligence Community had amassed a great deal of valuable intelligence regarding Usama Bin Ladin and his terrorist activities, none of it identified the time, place, and specific nature of the attacks that were planned for September 11, 2001. Nonetheless, the Community did have information that was clearly relevant to the September 11 attacks, particularly when considered for its collective significance."
what we infer: The load-bearing finding between proven negligence and unproven complicity, and it is routinely quoted in half. The first clause undermines specific foreknowledge; the second establishes that the relevant information existed and was not assembled. Both are the same finding.
limitations: Conducted while much of the underlying material remained classified, including the section not released until 2016.
Congressional bodies assessing agencies they oversee.

Senate Select Committee on Intelligence and House Permanent Select Committee on Intelligence (2002)AI-verified citationSEP-C018 SEP-C019

Undermines· strongSEP-E028

The inspector general did not substantiate the pre-attack identification

what the source states: "The investigation addressed nine specific allegations raised in the media and by various Members of Congress. We did not substantiate those allegations. The evidence did not support assertions that Able Danger identified the September 11, 2001, terrorists nearly a year before the attack, that Able Danger team members were prohibited from sharing information with law enforcement authorities, or that DoD officials reprised against LTC Shaffer for his disclosures regarding Able Danger." And: "We concluded that prior to September 11, 2001, Able Danger team members did not identify Mohammed Atta or any other 9/11 hijacker. While we interviewed four witnesses who claimed to have seen a chart depicting Mohammed Atta and possibly other terrorists or 'cells' involved in 9/11, we determined that their recollections were not accurate." The chart in question "was produced by Orion Scientific Corporation (Orion) in May 1999 and contained the names and/or photographs of 53 terrorists ... including a Brooklyn cell, but it did not identify Mohammed Atta."
what we infer: A specific documentary alternative to the witnesses' recollection: a real 1999 chart, with a real Brooklyn cell, that did not name Atta. The finding does not dispose of the dispute — the underlying data was destroyed — but it is the only reconstruction of the artifact anyone has published.
limitations: A department investigating allegations against its own senior officials.
Obtained from an archived capture of a mirrored copy of the official scan; no reachable government copy was located, and the quotes are OCR-derived.
This report describes four witnesses who recalled the chart; the separate figure of five concerns witnesses barred from testifying, and the counts must not be merged.

U.S. Department of Defense, Office of the Inspector General (2006)AI-verified citationSEP-C021

Undermines· strongSEP-E031

The Commission on Saudi funding — including the parenthetical usually dropped

what the source states: "Saudi Arabia has long been considered the primary source of al Qaeda funding, but we have found no evidence that the Saudi government as an institution or senior Saudi officials individually funded the organization. (This conclusion does not exclude the likelihood that charities with significant Saudi government sponsorship diverted funds to al Qaeda.)"
what we infer: The sentence is regularly quoted to close the question and regularly quoted, minus its parenthetical, to close it further than the Commission did. Read whole, it excludes institutional and senior-official funding while expressly leaving open diversion through state-sponsored charities — which is why SEP-C023 is unresolved rather than contradicted.
limitations: A 2004 finding; the declassification reviews ordered in 2021 and the subsequent litigation record postdate it.

National Commission on Terrorist Attacks Upon the United States (2004)AI-verified citationSEP-C023

Undermines· strongSEP-E036

The 'pull it' remark in its documentary context

what the source states: The remark as transcribed from the 2002 documentary: "I remember getting a call from the fire department commander, telling me that they were not sure they were gonna be able to contain the fire, and I said, 'You know we've had such terrible loss of life, maybe the smartest thing to do is pull it.' And they made that decision to pull and then we watched the building collapse." The spokesperson's account: "In the documentary, Silverstein's phrase 'pull it' meant pulling the contingent of fire fighters remaining in the building ... It had nothing to do with demolishing the building."
what we infer: The quotation is authentic and its immediate context is a conversation about whether the fire could be contained — which the contemporaneous record independently corroborates, since no manual firefighting was attempted for lack of water. The demolition reading also requires rigging a 47-story occupied building during an ongoing emergency, which is not feasible on that timescale.
limitations: The documentary itself was not viewed for this case, so no in-film timestamp locator exists; the quotation is taken from the fact-check that names the film.
The exculpatory reading comes from the speaker's own spokesperson — an interested party's account, not independent evidence of intent.

Sudiksha Kochi (2022)AI-verified citationSEP-C017

Undermines· moderateSEP-E005

The challenger study's conclusions, and its own hedge

what the source states: "The principal conclusion of our study is that fire did not cause the collapse of WTC 7 on 9/11, contrary to the conclusions of NIST and private engineering firms that studied the collapse. The secondary conclusion of our study is that the collapse of WTC 7 was a global failure involving the near-simultaneous failure of every column in the building." On its own limits: "We cannot completely rule out the possibility that an alternative scenario may have caused the observed collapse; however, the near-simultaneous failure of every column is the only scenario we identified that was capable of producing the observed behavior."
what we infer: The most substantial published challenge to the official mechanism, and more cautious in its own text than in its reception. Its stated conclusion is that it identified no fire-driven scenario reproducing the collapse — which is weaker than, and often reported as, a demonstration that none exists.
limitations: Funded by the advocacy organization campaigning for its conclusion, which the report acknowledges without stating the amount.
Not peer-reviewed in a journal.
Used a single static temperature distribution rather than a modeled fire progression.

J. Leroy Hulsey (2020)AI-verified citationSEP-C004 SEP-C003

Undermines· moderateSEP-E006

Technical criticism: one connection, one static temperature field, no fire progression

what the source states: "The study only focuses on one connection." "The study does not model fire progression. Dr. Hulsey only used one static temperature distribution, where the actual fires moved around heating unevenly." "The study neglects unknowns. Impact damage from falling WTC1 debris, the actual fire spread and temperatures, the state of the insulation at every spot, and differences between drawings and constructions are all factors that are unknown, and make it impossible make a determination of the exact cause of the collapse."
what we infer: These are methodological objections that do not require accepting the official mechanism to have force: a model driven by one static temperature field cannot represent the moving fires that the collapse hypothesis depends on. They bear on the challenger study's weight, not on whether the official model is right.
limitations: A discussion-forum post, not a peer-reviewed critique; its authority rests entirely on the arguments themselves.
The site is openly adversarial to the hypothesis under examination.

Mick West (2017)AI-verified citationSEP-C004

Undermines· moderateSEP-E007

Criticism: simultaneous column removal is the model's input, not its finding

what the source states: A participant's characterization of the method: "My idea of what they are doing to 'fail' the columns is this: They simply make all columns literally disappear. Then the building falls at free fall." And on the unaddressed evidence: "Hulsey failed to demonstrate that his best fit collapse progression model (Section 4.6) resulted in a debris distribution similar to Figure 1.7, and had large wall segements hitting Fiterman Hall and the Verizon Building."
what we infer: If near-simultaneous column failure is imposed rather than derived, the study demonstrates that its input produces the observed descent — not that the input occurred. This is the single strongest objection to SEP-C004 and the reason it is graded weakly supported.
limitations: Pseudonymous forum posts characterizing another team's method; the characterization has not been answered by the study's authors in a published document this case has obtained.

Mick West (2019)AI-verified citationSEP-C004

Qualifies

3 records · strongest first

Records that narrow, complicate, or bound what another record can carry.

Qualifies· moderateSEP-E010

The unexplained sulfidation, and the study its authors asked for and never got

what the source states: "Evidence of a severe high temperature corrosion attack on the steel, including oxidation and sulfidation with subsequent intergranular melting, was readily visible in the near-surface microstructure. A liquid eutectic mixture containing primarily iron, oxygen, and sulfur formed during this hot corrosion attack on the steel." And, on the open question: "The severe corrosion and subsequent erosion of Samples 1 and 2 are a very unusual event. No clear explanation for the source of the sulfur has been identified. The rate of corrosion is also unknown. It is possible that this is the result of long-term heating in the ground following the collapse of the buildings. It is also possible that the phenomenon started prior to collapse and accelerated the weakening of the steel structure. A detailed study into the mechanisms of this phenomenon is needed to determine what risk, if any, is presented to existing steel structures exposed to severe and long-burning fires."
what we infer: A genuine documented anomaly that no one has resolved in either direction, because the study its own authors requested has never been performed. Note what the appendix does NOT do: it proposes no source for the sulfur at all. The frequently repeated attribution of a gypsum explanation to these authors is mistaken; their stated alternatives concern when the corrosion occurred, not where the sulfur came from.
limitations: Two members recovered opportunistically, with hedged identifications ('appeared to be', 'thought to be') rather than documented provenance.
The appendix's own leading possibility is post-collapse heating in the debris pile, which would carry no implication for the collapse mechanism.

Jonathan Barnett (2002)AI-verified citationSEP-C007

Qualifies· moderateSEP-E016

The agency's stated reason for not testing steel for explosive or thermite residue

what the source states: Asked whether the steel was tested for explosives or thermite residues: "NIST did not test for the residue of these compounds in the steel." The reasoning offered: "Analysis of the WTC steel for the elements in thermite/thermate would not necessarily have been conclusive. The metal compounds also would have been present in the construction materials making up the WTC towers, and sulfur is present in the gypsum wallboard that was prevalent in the interior partitions."
what we infer: A passage both sides misdescribe in opposite directions, which is why it is recorded verbatim. "The investigation found no thermite" is false — it did not look. "Therefore thermite" is equally unsupported. The gypsum-as-sulfur-source argument is this agency's, made here; it is not the FEMA metallurgists' explanation, though it is frequently attributed to them.
limitations: The agency's justification for a methodological choice, not an independent assessment of whether the choice was sound.

National Institute of Standards and Technology (2021)AI-verified citationSEP-C014 SEP-C013

Qualifies· moderateSEP-E032

The 2025 ruling: jurisdiction, not merits — and now on appeal

what the source states: "A federal judge in New York on Thursday rejected Saudi Arabia's latest effort to dismiss civil claims that it supported the 9/11 hijackers." The judge "said in a written opinion that his decision pertained to jurisdiction rather than the merits of the claims against the Kingdom of Saudi Arabia." He "found the claims legally sufficient to proceed to trial, noting that an imam and an accountant's employment by Saudi Arabia likely had some connection with their support of two al-Qaida members who came to the United States in early 2000 to study English and take flight lessons."
what we infer: A finding of legal sufficiency, expressly not of responsibility. It raises the claim from stalled to triable and decides nothing about whether it is true. Current posture matters and this article predates it: the Kingdom filed a notice of appeal in September 2025 and proceedings were stayed, so nothing is presently proceeding to trial.
limitations: Journalism reporting a court order; the order itself supersedes this account and should be obtained.
The article carries no explicit statement of the Kingdom's denial, so none is attributed to it here.
Superseded on posture by later docket activity — the appeal and stay are not reflected in the article.

The Associated Press (2025)AI-verified citationSEP-C023

Context

4 records · strongest first

Background that informs the case without pushing either way.

Context· moderateSEP-E033

The declassification order and the release collection it produced

what the source states: "Executive Order 14040—Declassification Reviews of Certain Documents Concerning the Terrorist Attacks of September 11, 2001 / September 3, 2021," directing review of records "identified but withheld in full during discovery in In re Terrorist Attacks on September 11, 2001, MDL No. 03–1570 (S.D.N.Y.)."
what we infer: Establishes the mechanism by which the Saudi-file record has been moving since 2021, and independently confirms the litigation's docket from a government source. What the released records establish is a separate question this case does not yet grade, because the releases are heavily redacted and the characterizations in circulation are the parties' own.
limitations: An order about process; it establishes nothing about content.

Executive Office of the President (Joseph R. Biden, Jr.) (2021)AI-verified citationSEP-C022

Context· moderateSEP-E038

Baseline: 1,202 exhibits admitted in an adversarial criminal trial

what the source states: "The following web pages link to all 1,202 exhibits admitted into evidence during the trial of U.S. v. Moussaoui, with the exception of seven that are classified or otherwise remain under seal. This is the first criminal case for which a federal court has provided access to all exhibits online. The exhibits were posted on July 31, 2006."
what we infer: Establishes that a large body of the underlying evidence was tested in adversarial proceedings with defense counsel and a judge, and is publicly inspectable. This is why the authorship of the attacks is not a live dispute on this site, and why the disputes that remain live are the ones where no comparable adversarial test has occurred.
limitations: Seven exhibits remain sealed.
This case has not examined individual exhibits; the record is cited for the existence and scope of the adversarial test.

U.S. District Court for the Eastern District of Virginia (2006)AI-verified citationSEP-C019

Context· weakSEP-E034

The FBI release series under the 2021 order

what the source states: The Bureau's public records collection "9/11 Attacks Investigation and Related Materials" lists an Executive Order 14040 release series published in March and April 2022, including items designated responsive to subsections 2(b)(1), 2(c), and 2(d) with supplemental releases.
what we infer: Confirms that responsive records were released and are publicly available, which is the precondition for any claim about their contents. This case makes no claim about what they contain: the documents are heavily redacted, and every characterization of them currently in circulation comes from a litigant.
limitations: The live collection page is bot-walled; existence and contents-listing were verified from a dated archive snapshot rather than the live site.
No document in the series has been read for this case.

Federal Bureau of Investigation (2022)AI-verified citationSEP-C022

Context· weakSEP-E037

A broadcaster's account of its own premature report

what the source states: "We're not part of a conspiracy. Nobody told us what to say or do on September 11th. We didn't get told in advance that buildings were going to fall down. We didn't receive press releases or scripts in advance of events happening." And: "We no longer have the original tapes of our 9/11 coverage (for reasons of cock-up, not conspiracy)." And: "If we reported the building had collapsed before it had done so, it would have been an error - no more than that."
what we infer: Included as context for a widely circulated episode, and honestly limited: this is the interested party explaining itself, and the missing tapes are its own account of its own record-keeping. Note that the on-air time usually cited alongside this episode does not appear in this source and is asserted nowhere in this case.
limitations: A first-person institutional account of the institution's own error, with no independent verification of the explanation offered.
Contains no timestamps for either the report or the collapse.

Richard Porter (2007)AI-verified citationSEP-C017