The Department of War published its sixth PURSUE tranche on Friday, September 18, 2026, six weeks after the August 7 release. It is the longest gap yet and the second-largest batch, 75 records, behind only the opening tranche in May. The running corpus now stands at 450 released records across six tranches, and every one of them is indexed in the full PURSUE archive.
Forty-four of the 75 records belong to a single Defense Intelligence Agency contract from 2008. The rest split into four smaller sets: the Tremonton film of 1952 and its paperwork, a recording of the man who ran Project Blue Book explaining his plans, the first local police footage the archive has published, and the usual run of modern military infrared clips. Taken together, the tranche is less about sightings than about how the government has gone about studying them.
The contract behind the 2017 story
The Advanced Aerospace Weapon System Applications Program, AAWSAP, is the contract at the root of the Pentagon program the New York Times revealed in December 2017, then under the name AATIP, and that revelation set off the modern UAP era. The story has mostly been told from memory by people who worked on it. Release 06 publishes the paperwork, and the paperwork is narrower than the stories.
The July 2008 Statement of Objectives does not describe a UFO investigation. It asks a contractor to assess foreign aerospace threats out to 2050 by working out the physics of possible breakthrough technologies. It lists twelve technical areas, from lift and propulsion to “spatial or temporal translation”, signature reduction and human effects, and asks for research reports by the tenth month and a final integrated threat assessment. The September 2008 solicitation and original order turns that into priced line items and awards the work to Bigelow Aerospace Advanced Space Studies, the company set up by the hotel and space-habitat entrepreneur Robert Bigelow.
Five contract modifications follow, and together they show a program losing momentum. P00001 exercised the first option year in September 2009, but only “subject to the availability of appropriated funds.” P00002 funded that year in February 2010 while noting the “impacts to deliverables resulting from reduced Option Year 1 funding.” Two May 2010 changes swapped the government’s contract representative, and P00005 stretched the performance period from September 22 to December 21, 2010, with no new money. None of this settles what AAWSAP really was, but it shows a contract being cut back two years in.
The other 37 AAWSAP records are Defense Intelligence Reference Documents, or DIRDs. The list of 38 titles has been public since 2018 and several papers had surfaced through FOIA; Release 06 brings 37 together in one official collection, with a warning written into every summary: DIRDs “are best understood as reference and synthesis products rather than as original research,” and the summaries “should not be read as implying current validation of the concepts discussed.” They are literature reviews, and their range is wide. Some are routine surveys that would fit in any defense library: metallic glasses, ultracapacitors, high-energy lasers, and even biomaterials for medical implants. Others are the titles that made the program famous: warp drive and extra dimensions, traversable wormholes and “stargates”, antigravity, and energy from the quantum vacuum.
The summaries of those titles are the most striking writing in the release, because the government is assessing its own past work. The antigravity paper, it says, “deviates significantly from mainstream physics consensus.” The warp-drive paper “lacks empirical support and offers no viable engineering pathway.” The paper on anomalous field effects on human tissue, the one closest to encounter claims, relies on “case comparisons lacking physical examinations” and on claims “that lack independent scientific validation in the report.” The sharpest line is about the paper comparing lone “maverick” inventors with corporate labs. Its preferred innovator, the summary observes, looks a lot like the people who wrote the DIRDs, and so the paper shows “a notable methodological circularity, validating the program’s operational model rather than neutrally assessing the innovation ecosystem.”
That candour is why AAWSAP leads this briefing, though it contains no sightings. For nine years it has been cited as proof that the Pentagon took exotic physics seriously. The records show that DIA paid for surveys of exotic physics, and that the government now says in writing that the most exotic of them did not hold up. The AAWSAP hub page links every contract record and DIRD in the set.
Ruppelt, on tape
The most valuable historical item in the tranche is a recording. On March 26, 1952, in Boston, Captain Edward J. Ruppelt described the Air Force’s reorganized UFO investigation, the effort that became Project Blue Book. The release includes both the audio and a transcript. Ruppelt’s 1956 book, The Report on Unidentified Flying Objects, is still the best inside account of early Blue Book; this is him four years earlier, while the project was still being set up.
According to the government’s summary, of roughly 800 historical cases the Air Force had reviewed, about 20 percent remained unexplained after technical analysis, most of them reported by military personnel, airline pilots and other trained observers. He pointed to clusters of reports around sensitive nuclear sites, naming Los Alamos, Albuquerque and Oak Ridge. He described silent high-speed flight and radar returns suggesting speeds of 1,000 to 3,500 miles per hour, and the summary is careful to call those speeds unconfirmed. He also said plainly that the Air Force had found no conclusive evidence about what the objects were or where they came from.
What he proposed to do about it is the part that has held up best. Ruppelt wanted instruments: diffraction-grating photography to split a light into its spectrum, and radar scopes synchronized with cameras, so that a report would come with a physical measurement and not only a description. AARO makes the same argument today. The 20 percent figure is worth setting against the end of the story. When Blue Book closed in 1969 it had logged 12,618 sightings and listed 701 as unidentified, about 5.6 percent. The unexplained share fell as the files grew, a pattern repeated in almost every later dataset. His nuclear-site clusters echo the corpus’s wider nuclear pattern and Release 04’s Los Alamos green fireballs.
Tremonton: the film, the balloons, and the sewing machine
Three months after Ruppelt spoke, on July 2, 1952, Navy Chief Warrant Officer Delbert C. Newhouse stopped his car near Tremonton, Utah, and filmed a group of bright objects with a handheld Bell & Howell Auto Master camera. Release 05 carried the Navy’s 1953 frame-by-frame analysis of the footage. Release 06 adds everything around it: the Project Blue Book case file, a photo file on the “pillow balloon” review, Newhouse’s final Navy personnel record, a late-1952 Air Force “flying discs” file with eleven pages on the case, and the film itself. The full story is on the Tremonton case page.
The files show a witness everyone trusted and an explanation that kept changing. The Air Force judged Newhouse honest, reliable and expert, with more than 1,000 hours of aerial photography behind him, and his personnel record confirms he retired as a Chief Warrant Officer 4, the highest warrant grade of the time. In December 1952, Blue Book staff, balloon specialists and representatives of General Mills (then the only maker of pillow balloons) watched the film together. Most agreed the objects looked and moved like pillow balloons. One General Mills man disagreed, saying they seemed too bright and moved too fast. Earlier speed calculations, the group noted, had assumed Newhouse held the camera perfectly still. No balloon launch in the area was ever found, and later Air Force reviews settled on seabirds reflecting sunlight. With no fixed reference points, distance, size and speed cannot be measured from the film, a limit that applies to the seabird answer as much as any other.
Then there is the sewing machine. From the 55-second mark, the digitized print held by the National Archives switches to unrelated footage of what AARO calls “a demonstration of sewing equipment.” Newhouse’s own paperwork says he used the rest of his 16mm roll on mountain scenery in Idaho, and that footage is gone. The surviving print looks like a composite or a reused reel, and, as the government says, the record “does not establish how or when the materials came to be combined.” One of the two most famous UFO films of the 1950s survives in the national archive spliced onto an industrial film, which says a good deal about how this material was kept.
The first police officers in the archive
Eight records in this tranche use a new prefix, LLE, for local law enforcement. They are the first police submissions PURSUE has published. Colorado officers filmed on their phones four times: three videos from October 2023 (1 minute 19 seconds, 15 minutes, and 27 seconds) and one from January 2024 running 1 minute 42 seconds. Each has a matching transcript (D001, D002, D003, D004). AARO stripped the audio, blurred the lower half of each frame and redacted the transcripts.
All four reports describe the same thing: a light that was “silent” and showed “red, blue, and green flashing lights.” In the longest clip the camera holds one light near the center of the frame for a full quarter of an hour while the officer zooms in and out.
That description also fits the most common cause of police UFO calls: a bright star low on the horizon. Near the horizon, starlight crosses so much turbulent air that it twinkles hard and splits into flickering red, green and blue. It is silent, barely moves in fifteen minutes, and at full phone zoom swells into a blob that seems to pulse and change shape. Capella and Sirius cause this kind of report every autumn and winter in the northern hemisphere. That is not a verdict: nothing released gives the direction, elevation or time needed to check, and details in the transcripts may rule it out. But it is the first explanation to eliminate, and AARO’s own caveat points the same way. AARO attached the same note to all four videos: “Cellular devices of the type used to capture this footage employ autofocus and other automated image-processing functions that may distort the apparent morphological and kinematic characteristics of distant subjects, particularly at high magnification.” In the January clip, headlights from approaching cars twice throw the phone’s autofocus off during recording.
One small data error: the January 2024 video’s metadata gives its incident date as October 2023.
The modern military set
The rest of the tranche is sensor material from U.S. Central Command and Indo-Pacific Command, several clips paired with the mission reports (MISREPs) that describe them. The Iraq 2022 mission report describes objects “flying sporadically within the target area… at various speeds between 80 and 180 miles per hour,” one to two meters across, at an altitude the operator could not determine. Two infrared stills go with it (PR130, PR131). AARO notes the trail on each “is consistent with known image-compression and processing effects and does not necessarily represent a physical feature.” A 2025 Middle East report describes something that looked “like a white round orbit… moving as fast as 500 miles per hour,” with a five-minute clip in which the sensor loses and finds the object several times and its auto-tracker never locks on. A second 2025 report describes “six small spherical objects grouped together,” agile and “frequently changing direction,” at an estimated 480 miles per hour, with one minute of video.
Be careful comparing records here. The government’s summary of the 2022 Middle East mission report uses the “six small spherical objects” and 480 mph wording from the 2025 report almost word for word. Yet the two videos it accompanies quote that same 2022 report as giving 80 to 180 miles per hour. One of those summaries is wrong, and the release does not say which.
Six short infrared clips from 2023 finish the set: five over the Yellow Sea (PR143, PR144, PR150, PR151, PR152) and one over the East China Sea. They run from 19 seconds to just over a minute and a half, and none comes with a mission report. AARO’s descriptions amount to a sensor panning to keep a patch of contrast centered, which is what most of the modern corpus looks like.
What Release 06 adds up to
This tranche has no standout document like the Bahia cables; its value is in what it shows about method. Ruppelt in 1952 said the problem could only be solved with instruments. AAWSAP in 2008 paid for theory instead and got back papers that the government now marks as unsupported. The Tremonton film survives as a copy spliced onto a sewing demonstration. The first police evidence in the archive is phone video of a twinkling light, which comes with the government’s own caution about phone cameras. None of this shows that nothing is there. It does show that 74 years after Ruppelt asked for better data, the archive still mostly lacks it. The program overview is at /pursue/, and every document page is indexed at /disclosure/archive/.