UAP for India
27 July 2026 — Analysis — Explainer
Read as a series rather than as four separate documents, the Pentagon’s annual UAP reports show an archive that has nearly quadrupled since 2022 while the residue of cases nobody can explain has fallen from a wide-open question to a handful of single-digit files — even as a presidentially-ordered declassification sprint through 2026 has, for the first time, started to complicate that tidy trend.
On 20 July 2026, the Pentagon’s All-domain Anomaly Resolution Office quietly posted its Fiscal Year 2025 Consolidated Annual Report on Unidentified Anomalous Phenomena — a week old as of this writing, and the newest entry in a series that began in October 2023. The document’s masthead now reads “Department of War” rather than “Department of Defense,” following President Trump’s September 2025 renaming order.[1]
Its substance, taken alone, looks familiar: 319 new reports, a cumulative archive of 1,870 cases, and nine still flagged as needing further analysis. But this report no longer arrives in isolation. Since February 2026 it has been overtaken, in practical terms, by a separate presidentially-ordered declassification programme that is releasing hundreds of historical files on a rolling basis, has already produced at least one internal AARO document contradicting the annual report’s tidy resolution arithmetic, and has left a congressionally-mandated second volume of AARO’s own historical review still unpublished.
Read together — four annual reports, a historical review, and a still-unfolding disclosure sprint — the record shows a case count climbing steadily, a resolution process getting measurably better at explaining what it sees, and a gap between what AARO publishes on its own schedule and what Congress and the White House are now demanding that has, if anything, widened rather than closed.
The All-domain Anomaly Resolution Office, known as AARO, was created inside the Office of the Secretary of Defense on 20 July 2022, when the Deputy Secretary of Defense issued a memorandum expanding and renaming its predecessor, the Airborne Object Identification and Management Synchronization Group.[2] Its reporting duty predates the office itself: the first annual assessment, the Preliminary Assessment: Unidentified Aerial Phenomena of 25 June 2021, was published directly by the Office of the Director of National Intelligence and catalogued 144 reports, mostly from military aviators between 2004 and 2021, of which it could offer a firm explanation for exactly one.[3]
The office’s legal foundation is codified at 50 U.S.C. § 3373, which directs the Secretary of Defense and the Director of National Intelligence to jointly run a mechanism for detecting, identifying and, where possible, attributing unidentified anomalous phenomena across air, sea, space and — the novel addition — “transmedium” objects that appear to cross between those domains, with an annual report due to Congress by 31 October each year.[4] A reader who wants the fuller terminology this statute formalised should see our companion piece on what “UAP” actually means and why it replaced “UFO”.
The 2022 Annual Report on Unidentified Aerial Phenomena, published by ODNI in January 2023 and covering the year to 30 August 2022, added 247 newly reported sightings plus 119 older cases surfaced after the 2021 assessment, taking the cumulative total to 510.[5] Reporting at the time traced much of that jump to the office’s newly opened internal reporting channels driving the count up, rather than to any real change in underlying aerial activity — a chronic collection bias that recurs throughout this series, since nearly all usable sensor data still comes from military aviation, the one community with a habitual reporting culture.[6]
The build-out was not smooth: in February 2023, a bipartisan group of senators led by Mark Warner and Tim Kaine warned the Deputy Secretary of Defense and the Principal Deputy Director of National Intelligence that the fledgling office faced a funding shortfall threatening its science and technology plan, and pressed the Pentagon to reprogram money toward it.[7] The UAP Program Report User Guide that followed formally defined “UAP material” and “UAP engagement” as terms of art for the first time.[8]
That vocabulary underpinned the Fiscal Year 2023 Consolidated Annual Report on UAP, released 25 October 2023 and covering 31 August 2022 to 30 April 2023, which logged 291 new reports — 274 from the window itself, 17 backlogged from 2019–2022 — pushing the archive to 801. Over 100 of those new reports came through the Federal Aviation Administration, describing unshaped lights between 5,000 and 60,000 feet, with the share above 40,000 feet rising from 6 to 10 percent of the dataset.[9]
The Fiscal Year 2024 Consolidated Annual Report, released 14 November 2024 and covering 1 May 2023 to 1 June 2024, added 757 new reports — 485 current, 272 backlogged from 2021–2022 — taking the archive to 1,652 cases as of 24 October 2024. Of the 757, AARO had resolved 292 by publication, mostly balloons, birds and small drones, with 21 cases assessed as needing further analysis.[10]
Five days later, in testimony to the Senate Armed Services Subcommittee on Emerging Threats and Capabilities on 19 November 2024, incoming director Jon Kosloski put a figure on the residue directly: fewer than 3.5 percent of cases carried any signature suggestive of genuinely novel technology, though he added that a handful of “particularly interesting cases” remained ones he personally could not explain.[11] Kosloski had taken over as director on 24 August 2024, arriving from the National Security Agency’s Research Directorate after founding director Sean Kirkpatrick’s December 2023 departure and an interim period under Timothy Phillips.[12]
The gap between publication cycles is where this series gets easiest to misread. By 25 February 2026, more than a year after the FY2024 report and nearly five months before the FY2025 report would appear, Pentagon spokesperson Sue Gough told reporters that AARO was “examining over 2,000 UAP cases” — roughly 350 more than the FY2024 report’s already-dated 1,652, with around 1,000 of those retained in an active archive for lack of sufficient data to resolve them.[13]
That figure was already loose by the time the Fiscal Year 2025 Consolidated Annual Report itself finally appeared on 20 July 2026: covering 2 June 2024 to 30 May 2025, it formally recorded only 319 new reports for that window — a sharp drop from FY2024’s 757, which AARO attributes to having worked through a backlog of older FAA data in the previous cycle — and a cumulative archive of 1,870 cases as of its own 30 May 2025 cutoff, a figure already roughly 14 months stale against Gough’s February 2026 comment by the time the public read it.[14]
Of the 319, 114 were resolved during the period and another 256 older cases were closed retrospectively, for 370 total resolutions; 191 remain unresolved for lack of data. The anomalous residual fell again on paper, to nine cases referred for further analysis, down from 21 the year before — though the report also recorded a single maritime-domain case for the whole year: roughly one hundred airborne objects and two likely uncrewed surface systems that US Navy assets tracked off the Virginia coast, an incident still under active investigation as this piece is being written.[15] Independent trackers noted the report arrived several months past the reporting cadence Congress had come to expect, continuing a pattern of slippage across every cycle in this series.[16]
The gap between publication cycles is where this series gets easiest to misread.
Political scientists use the term institutional capacity to describe a state’s demonstrated ability to build durable, specialised bureaucratic machinery around a problem, as distinct from simply reacting to it case by case. Read this way, AARO’s reports are not primarily a scientific record of anomalous sightings; they are evidence of a capacity-building project — a dedicated office, a standing legal mandate, a growing sensor and modelling toolkit, and a repeatable output — that exists whether or not any individual case turns out to be interesting.
One independent assessment of the FY2025 data made this point precisely: better analysis tends to move cases out of the unidentified category rather than reveal a growing set of objects with demonstrably extraordinary performance, which means the trend the reports are actually documenting is a measurement problem being solved, not a phenomenon being discovered.[17] Every AARO report to date has recorded zero confirmed instances of the “foreign adversary systems” category, itself a capacity claim: an assertion that the office’s sensors and analytic tools are now good enough to rule that possibility out with confidence, not merely fail to find it.
In international relations, signalling refers to a visible or costly action a government takes primarily to communicate resolve or responsiveness to a particular audience, rather than to solve the underlying problem the action nominally addresses. The statutory deadline of 31 October has been missed in every cycle of this series, and Congress has responded not by relaxing the requirement but by widening it.
The Fiscal Year 2026 National Defense Authorization Act, signed into law on 18 December 2025 as Public Law 119-60, added three provisions in its Subtitle E amending 50 U.S.C. § 3373 to require AARO to brief Congress on the number, location and nature of UAP intercepts conducted by NORAD and US Northern Command going back to 1 January 2004, along with the procedures followed and any data collected.[18]
That was, notably, the narrower of two options on the table. A more sweeping amendment — SA 3111, introduced by Senators Schumer, Rounds and Gillibrand, which would have created a presumption of disclosure for UAP records and given an independent review board eminent-domain power over private contractors’ UAP materials, modelled on the JFK Assassination Records Act — was stripped from the Senate’s NDAA package on 9 October 2025 before conference negotiations even began, the third consecutive year a comprehensive disclosure framework has been narrowed or blocked during the legislative process.[19]
That specific claim about Pentagon-adjacent opposition rests on secondary reporting rather than a leaked primary document, and should be read as attributed analysis rather than an established fact; our companion piece on the record of congressional disclosure efforts traces the pattern across all three years in more depth. What Congress keeps getting, in other words, is a narrower reporting obligation rather than the disclosure architecture advocates have asked for — which is itself the gap this article’s title refers to.
What Congress keeps getting is a narrower reporting obligation, not the disclosure architecture advocates have asked for.
Epistemic authority is the socially recognised right to be believed on a factual question — here, the right to say definitively what a given sighting was. AARO’s five-category resolution taxonomy, examined below, is the mechanism by which the office holds that authority: it alone decides which cases move from “unresolved” into an explained bucket, and which stay in the small residual category the press then reports as “truly anomalous.” That authority is actively managed, not simply exercised.
Internal emails released under the Freedom of Information Act showed Pentagon public-affairs staff deliberately minimising the Office of the Director of National Intelligence’s statutory co-authorship role when messaging AARO’s Historical Record Report to the press, even though the underlying reporting requirement runs jointly to both agencies.[20] Deciding who gets credited as the source of a finding is itself an exercise of epistemic authority, and it recurs in 2026 in a different form, discussed in the final section: a single declassified file, signed by the same director who testifies to the “3.5 percent” figure, that reports a far higher unresolved rate for one specific set of sightings.
Across all four annual reports, AARO has used a consistent five-category framework. Airborne clutter covers birds, balloons, hobbyist drones and windblown debris that muddy a sensor picture. Natural atmospheric phenomena covers ice crystals, moisture and thermal gradients that register falsely on infrared or radar. US government or industry developmental programmes covers classified or proprietary domestic aircraft and sensor platforms an operator simply was not briefed on. Foreign adversary systems covers technology fielded by China, Russia or another state or non-state actor — the category every report to date has found zero confirmed instances of.
And “other” is the catch-all for cases that resist attribution, whether from thin sensor data, imperfect witness recall, or something that genuinely does not fit the known categories; it is the shrinking tail of “other” cases — 21 in FY2024, nine in FY2025 — that gets reported as “truly anomalous,” and it is worth being precise that the label signals unresolved, not unexplainable.
The label signals unresolved, not unexplainable.
The clearest mechanical upgrade behind the shrinking residue arrived in FY2025: newly deployed three-dimensional modelling and simulation tools let AARO attribute 238 previously ambiguous reports to satellite flaring — sunlight reflecting off orbiting hardware — and resolve all 44 space-domain cases in the reporting year to the same cause, a capability the office did not have in earlier cycles.[21] AARO’s own information paper on the technique explains the underlying method: matching a witness’s reported time, bearing and elevation against known satellite orbital data to reconstruct whether a bright, slow-moving light was in fact a glint off a solar panel.[22]
Separate from the annual case-count reports, AARO published a one-off document on 8 March 2024: the Historical Record Report, Volume I, covering 1945 through 31 October 2023 and mandated by Congress specifically to examine decades-old claims of a secret US crash-retrieval and reverse-engineering programme — a different document, with a different scope, from the FY2023 case-count report released four and a half months earlier. AARO’s conclusion was blunt: after interviewing more than 30 current and former officials and reviewing decades of records, the office said it had found no evidence of extraterrestrial reverse-engineering, and it specifically assessed a piece of recovered metal, long claimed to be of off-world origin, as terrestrial and unremarkable.[23]
A second volume was mandated in the same legislation, under Section 6802 of the FY2023 NDAA, to cover material and interviews not yet secured by the time Volume I was published. As of the most recent public accounting of AARO’s releases, dated 21 July 2026, that second volume still had not appeared — a gap former deputy assistant secretary of defense Christopher Mellon pointed to directly in February 2026 when he noted that AARO had “yet to fulfil its statutory obligations,” citing both the missing Volume II and, at that point, the still-unpublished FY2025 annual report.[24]
The FY2024 report’s publication in November 2024 was also Kosloski’s first formal appearance before Congress as director, and the substance of the message he delivered — a hard number for the anomalous share of cases, paired with a personal admission that some individual cases still puzzled him — illustrates how a leadership handover doubles as a credibility exercise. An NSA research pedigree lent the office technical authority precisely at the moment its predecessor’s December 2023 departure could have read as instability; the substance of the resolution framework did not change, but who was vouching for it did.
That is path dependency in miniature — the way earlier institutional choices constrain what a later actor can plausibly do, even after circumstances change: once an institution has committed to a taxonomy, a reporting cadence and a public position, each subsequent leader inherits and restates that architecture rather than rebuilding it, which is part of why four reports under two directors and one department rename still read as a single continuous document.
Strip away the headlines and the four-report series shows three consistent movements. First, the absolute case count keeps climbing — 510, then 801, then 1,652, then a reported 1,870 as of May 2025, north of 2,000 by informal count in February 2026 — largely because AARO keeps opening new reporting channels rather than because sightings themselves are accelerating. Second, resolution is visibly improving: the FY2025 report’s ability to model every space-domain case to satellite flare with new three-dimensional tools is the clearest evidence that better instruments, not more dramatic phenomena, explain a meaningful share of the “unexplained” pool from earlier years. Third, the truly anomalous residual has fallen even as the total case count has risen — from 21 flagged cases in FY2024 to nine in FY2025.
The truly anomalous residual has fallen even as the total case count has risen.
But 2026 produced a data point that cuts against that clean trend: a declassified AARO file dated 5 June 2026 and signed by Kosloski, covering a two-day flap of sightings in October 2023 that included an orange “mother” orb reportedly launching smaller red orbs, stated that roughly 40 percent of the reported phenomena in that specific file lacked a reasonable explanation.[25]
That figure describes one narrow batch of historical sightings under review through the disclosure programme discussed below, not AARO’s annual aggregate resolution rate, and the two numbers should not be read as contradicting each other so much as measuring different populations of cases — but it is a reminder that the tidy downward trend in the annual reports is not the whole picture once older, previously unreviewed files start being opened up case by case.
The starkest gap in this series sits between AARO’s written conclusions and sworn testimony given directly to Congress. On 26 July 2023, the House Oversight Committee’s National Security Subcommittee held a nationally televised hearing at which former intelligence officer David Grusch testified under oath that the United States has run a multi-decade UAP crash-retrieval and reverse-engineering programme, and that he had spoken with officials who claimed direct knowledge of recovered non-human biologics; in the same hearing, Navy pilots Ryan Graves and David Fravor described years of unexplained radar and visual contacts, including formations of dark grey cubes inside a clear sphere tracked off the US Atlantic coast.[26]
AARO’s position, restated in every report since and made explicit in the Historical Record Report, is close to the opposite: the office says it has found no verifiable evidence to substantiate claims that a programme to possess or reverse-engineer extraterrestrial technology has ever existed. Kirkpatrick, while still director, said publicly that the central source behind the crash-retrieval allegations — understood to mean Grusch — had declined to speak directly with AARO’s investigators, a claim Grusch disputed.[27]
A subsequent analysis of the 2023 hearings concluded that the testimony, however striking, did not itself shift the government’s evidentiary record.[28] Three years on, neither side’s account has been resolved by anything published since.
What has changed since the FY2024 report is that the annual-report cadence stopped being the only, or even the main, channel through which AARO-related material reaches the public. On 20 February 2026, President Trump posted that he would direct Secretary of War Pete Hegseth and other agencies to “begin the process of identifying and releasing Government files related to alien and extraterrestrial life,” a pledge that followed viral podcast comments by former President Barack Obama about the subject.[24] Five days later, Hegseth committed the Pentagon to “full compliance” with the resulting executive order while describing the effort as “a deliberative process” with no fixed timeline.[13]
The operational result was the Presidential Unsealing and Reporting System for UAP Encounters, or PURSUE, a Department of War platform at war.gov/ufo publishing declassified UAP-related files on a rolling basis: an initial tranche of roughly 160 records — reports, sensor videos, photographs and witness accounts spanning 1944 to the present — on 8 May 2026, a second batch on 22 May, a third on 12 June that reportedly included an FBI file tranche, and a fourth on 10 July 2026.[29]
Alongside PURSUE, a new UAP Governance Board spanning the Office of the Director of National Intelligence, the FBI and the Department of War stood up a UAP Science Advisory Council in mid-June 2026, led by Harvard astrophysicist Avi Loeb and drawing on specialists from oceanography, molecular biology, psychology and instrumentation, with its first meeting held on 18 June 2026.[30]
None of this replaces the statutory annual report, and none of it produced Volume II of the Historical Record Report; if anything, it has made AARO’s own publication schedule look slower and thinner by comparison, precisely the gap between what the office is legally required to publish and what its political leadership is now promising that this series exists to track.
AARO is required to keep publishing this annual series, and every edition to date has arrived later than the one before it — October, then November, then July of the following year. That is the office’s own path dependency working against it: once a reporting cycle slips its statutory deadline once, later slippage becomes easier to normalise rather than harder.
Congress’s response, in the FY2026 NDAA, was to add reporting obligations rather than tighten the deadline, which means the test for the next cycle is not whether AARO publishes on time but whether the NORAD and Northern Command intercept data now required by law actually appears, whether Volume II of the Historical Record Report finally surfaces, and whether the PURSUE disclosure programme keeps running independently of AARO’s own reporting cadence or is eventually folded back into it. Readers wanting a fuller treatment of what the accumulated evidence can and cannot support should see our analysis of why UAP is now a national security problem rather than a curiosity.
What can be said with confidence, on the paperwork alone and as of 27 July 2026, is this: across four annual reports and a parallel declassification sprint, the archive has grown well past its own most recent official count, the explained share of it has grown faster still, and the small number of cases nobody in the Pentagon can currently explain has been shrinking in the aggregate even as at least one newly opened historical file shows a far higher unresolved rate — while the most serious allegations of a cover-up remain exactly where they were in July 2023, unresolved by any report, release or hearing since. If you have a first-hand UAP sighting relevant to India, you can submit it through our report form.