On July 7, 2012, police in Cyprus arrested a man named Hossam Taleb Yaacoub. He was a Hezbollah operative, and he had been doing what terrorist operatives do before an attack: watching. He had recorded the arrival times of Israeli Arkia flights, the registration numbers of tour buses, the security arrangements at hotels, how close the emergency services were. Cypriot surveillance teams watched him watching, followed the pattern of his pattern-collection, and arrested him before the pattern became a bombing. He was convicted in March 2013 [9].
Eleven days after his arrest — July 18, 2012 — the same tradecraft ran undetected at Burgas airport in Bulgaria. A bomber who had blended into the flow of an ordinary tourist airport boarded a bus full of Israeli tourists and detonated. Five Israelis and the Bulgarian driver were killed [9]. Same organization, same surveillance method, same target set, eleven days apart. In one country the pattern was seen in time. In the other it was not. That difference — seen or unseen — is the whole mechanics of terrorism, and it is the subject of this article: terrorism is a patterns business, counterterrorism is a pattern-recognition business, and a particular kind of mind, the autistic mind, is built for the second in a way the rest of us are not. The evidence for that claim is specific, and this article will keep it specific — including where it runs out.
The attacker’s side: patterns as camouflage
Start with what the attackers actually do, because it is not hiding. It is rehearsing in plain sight, inside the pattern of ordinary life, counting on the pattern to make them invisible. The cleanest documented case is Mumbai. Between September 2006 and July 2008, David Coleman Headley made five surveillance trips to India under the cover of an immigration business, videotaping the Taj Mahal Palace hotel, the Oberoi, Leopold Café, the Chhatrapati Shivaji terminus, and the Chabad house, and taking boat trips to scout the landing sites he marked with GPS — reporting back to his handlers in Pakistan after every trip. His federal indictment reads like a syllabus [5]. In November 2008, ten gunmen who had studied his product killed about 164 people at the targets he had filmed. Headley was not invisible because he was careful once. He was invisible because he did an ordinary thing — a businessman with a camera, five times, over two years — and ordinary things are filed by the human mind under “normal” and never looked at again.
The pattern repeats across the decades. Al-Qaeda’s surveillance team photographed and diagrammed the U.S. embassy in Nairobi from December 1993; on the surveillance photo, Osama bin Laden himself pointed to the spot where a truck could be parked — and in August 1998 a truck bomb exploded at roughly that spot [5]. The London bombers of July 7, 2005 ran a full dry run on June 28 — Luton to King’s Cross to Baker Street, three hours, mingling with commuters — which the Metropolitan Police’s Peter Clarke identified afterward, flatly: “It was a dry run” [5]. Al-Shabaab allegedly rented a shop inside Nairobi’s Westgate mall weeks before the September 2013 attack and used the service lifts to stockpile weapons behind the pattern of retail [5]. And before October 7, 2023, Hamas spent at least seven years collecting on the Gaza border towns — a November 2020 document from captured computers catalogued the IP addresses and serial numbers of the security cameras of entire regional councils, down to Kfar Aza’s cameras and its guards’ phone lists [5]. U.S. intelligence’s own guidance to first responders describes the method in so many words: preoperational surveillance builds familiarity with layouts, security posture, and “persons’ patterns of life,” using blending tactics to do it [5].
Understand what the camouflage is made of. It is not darkness or disguise. It is routine — yours. The operative’s cover is the observer’s filing system: the part of the brain that sees a thing the tenth time and stops seeing it at all.
Why ordinary seeing fails
The laboratory version of this is famous. In Simons and Chabris’s 1999 experiment, subjects counting basketball passes missed, about half of them, a person in a gorilla suit who walked into the middle of the game, stopped, faced the camera, and thumped her chest [10]. The gorilla was not hidden. She was filtered — by attention organized around a schema, a task, a story about what the scene contains. Psychologists call it inattentional blindness, and its field version is deadlier than the lab’s. In the months before October 7, Israel’s border observers — the tatzpitaniyot, mostly young women who watched the Gaza fence for a living — reported Hamas rehearsing on a replica tank, a mock post, a mock fence: the attack, practiced in the open. Observer Noa Melman said afterward that “everyone treated it like it was normal, like it was routine.” Sergeant Roni Eshel, killed at Nahal Oz, had reported broken cameras and holes in the fence [10]. The eyes worked. The reports were filed. What failed was the schema above them — the governing concept, the conceptzia, that Hamas was deterred, which reclassified every anomaly as routine. Terrorism does not need your eyes to fail. It needs your filing system to succeed.
The autistic profile: what the science actually shows
Against that background, here is the peer-reviewed record on autistic perception — stated at its real strength, which is narrower than the mythology and stronger than the dismissal. The most replicated autistic strength in the literature is performance on the Embedded Figures Test: finding the hidden shape inside a complex design, the pure form of “the detail that does not belong.” In visual-search experiments (Joseph and colleagues, 2009), autistic subjects were faster overall, and the mechanism matters: the advantage lay in target–distractor discrimination at the point of attention — seeing, at the spot being examined, that this is not that — rather than in search strategy [1]. Most relevant to surveillance work, two laboratory studies found reduced inattentional blindness under load: typical adults’ detection of an unexpected stimulus collapsed as perceptual load rose; autistic participants kept detecting it, with no cost to their primary task (Remington, Swettenham and Lavie, 2012; Swettenham and colleagues, 2014) [1]. The theoretical frames have names — Baron-Cohen’s systemizing, the drive to find the lawful pattern in data; Mottron’s enhanced perceptual functioning — but the operational translation is simpler: a perceptual system less governed by schema and more by the data itself is, on tasks with a definable baseline, less likely to file the anomaly away [1].
Now the limits, because this article does not print the superpower version. A 2015 meta-analysis in Psychological Bulletin — 56 studies, roughly a thousand autistic participants (Van der Hallen and colleagues) — found no overall enhancement of local processing and no global deficit, only slower global processing; the advantage is real on some cluttered visual tasks and does not generalize across the board [2]. The spectrum is heterogeneous; savant-level ability is a minority phenomenon; change-blindness results are mixed and mostly from adolescents. The accurate claim is a profile, not a cape: task-specific strengths in detail discrimination, sustained focus on data streams, and resistance to schema-driven filtering — exactly the profile of imagery and data analysis, and not of, say, reading a source’s face across a café table.
Israel built the pipeline; the West is following it
One country stopped debating the profile and started staffing to it. In 2013 the IDF-affiliated programme Ro’im Rachok (“Seeing Far”) began taking autistic Israelis — normally exempt from service — as volunteers: three months of professional and life-skills training at Ono Academic College, three to four months provisionally inside a unit, then a volunteer draft. The first course trained its soldiers to decipher aerial and satellite photographs for Unit 9900, Israel’s visual-intelligence unit, on the programme’s own stated rationale: they are “very visually-oriented… patient and have the ability to focus on details” [3]. The programme has since expanded to more than ten units — software quality assurance, information sorting, electro-optics, electronics, the Air Force [3]. The testimony is programme-level but it is consistent: a Unit 9900 imagery decipherer, Corporal O., describing spotting the same object across frames — “I’d locate it in every picture — six pictures in a row”; the IDF’s recruitment chief, Lt. Col. Rotem Sabag, in December 2023, calling it a “qualitative advantage… the ability to concentrate, remember specific details” [3]. A fundraising organization’s page credits soldiers on the spectrum with a significant role in the imagery work of Operation Breaking Dawn in August 2022; that is advocacy-grade sourcing and is labelled as such here [3].
The honest gap must be stated in the same breath: no public source names a specific autistic analyst detecting a specific attack, tunnel, or plot. Unit 9900’s work is classified; the programme’s published record is placements, commanders’ assessments, and retention — about 80 percent of roughly a hundred applicants a year accepted, with the programme candid that candidates must be high-functioning and self-sufficient. That is a selection effect, and it means the results describe selected autistic soldiers in matched roles, not “autism” in general [3]. What can be said without inflation is that a serious military, fighting for its life on a border where pattern failure killed 1,200 people, looked at the evidence and institutionalized the bet: for the imagery chair, it recruits the pattern-seeing mind on purpose.
The Western agencies have made the same bet more quietly. The U.S. National Geospatial-Intelligence Agency — the American analogue of Unit 9900 — launched a neurodiversity pilot in December 2020 with MITRE and Melwood, placing interns in geospatial and imagery analysis; its deputy director, Dr. Stacey Dixon, said neurodiverse talent “adds significant value to the geospatial-intelligence tradecraft” [4]. At GCHQ, an autistic staff member writing officially under the name James described the job fit in 2019: “Very good attention to detail, can spot patterns, anomalies and trends easily… crucial to GCHQ’s work” — noting a high number of neurodiverse staff “stretching back to Bletchley Park” and a specialised neurodiversity support service running for twenty years [4]. A RAND study of neurodiversity in the intelligence community records the strengths the agencies cite — pattern recognition, visualization, focus — and carries the community’s favourite anecdote, which is printed here as an anecdote: an employer who “bragged that his autistic employees could look at a blurry satellite image obscured with foliage and tell the difference between a Russian MiG, a Ukrainian MiG, and a Russian MiG painted like a Ukrainian MiG” [4]. The United States also exhibits the contradiction in full: the Defense Department excludes autistic people from uniformed service entirely, while its civilian intelligence agencies recruit them — the left hand barring the mind the right hand is hiring [4].
The saves on record all share the profile
Now return to the documented saves, and notice what they have in common — because it is the profile above, worn by people who were never tested for anything. At Heathrow on April 17, 1986, El Al security questioned Anne-Marie Murphy, the unwitting, pregnant courier of the bomber Nezar Hindawi; her story was incoherent, but the physical catch was simpler: her bag, emptied, was unexpectedly heavy — a deviation from what an empty bag weighs, noticed by people who handled bags all day. Inside: about a kilogram and a half of Semtex in a false bottom, wired to a calculator, meant for El Al Flight 016 with some 395 people aboard [7]. In Times Square on May 1, 2010, the men who flagged Faisal Shahzad’s smoking Pathfinder were two street vendors, Duane Jackson and Lance Orton — men who worked that block every day, for whom the block was a baseline, and the wrongly parked SUV with its engine running and keys in the ignition was a deviation that popped [7]. In Cyprus, the save was surveillance of surveillance [9]. And in Abbottabad, the case against the compound was a list of deviations from baseline: eighteen-foot walls and barbed wire where they did not belong, no telephone or internet, rubbish burned instead of collected, no visible income, and a tall man who paced but never left — until 1–2 May 2011. The deepest irony in the file: bin Laden’s excessive pattern-avoidance was itself the pattern [8].
A bag that weighs wrong. A car parked wrong on a block a man knows by heart. A compound too quiet for its size. None of these observers was a genius, and none was working from a theory. Each had a baseline so thoroughly learned that the anomaly announced itself. That is the entire job description of the imagery chair and the data desk — a baseline, learned cold, and a mind that notices the deviation instead of filing it. It is also, on the laboratory evidence, the cognitive situation in which the autistic profile’s documented strengths live: discrimination at the point of attention, pattern held against pattern, and a filing system that does not rush to call the strange thing routine. The programmes in Jerusalem and Virginia are not charity and not symbolism. They are an attempt to hire, at scale and on purpose, the one cognitive profile the saves keep exhibiting by accident.
The limits, stated — and the real lesson
Two cautions complete the record rather than undermine it. First, unstructured human behaviour in crowds is a different task from structured imagery and data, and nobody is reliably good at it: the U.S. Government Accountability Office found in 2013 that behaviour-based deception detection — the basis of the TSA’s SPOT programme, some $900 million spent since 2007 — performed the same as or slightly better than chance, about 54 percent, across four meta-analyses of more than 400 studies, and recommended limiting its funding [6]. The autistic advantage, where it is documented, lives where a baseline can be defined — pixels, data streams, schedules, compounds — not in reading strangers’ faces in a terminal. Second, every programme in this article assigns its autistic personnel to technical chairs — imagery, quality assurance, data — and none, on the public record, to source handling, interrogation, or field surveillance-detection, the social-inference roles the same literature flags as the hardest for this population. That is not a failure of the thesis. It is the thesis, applied with discipline: match the mind to the task where its pattern is an advantage, and do not ask it to be a different mind.
Which leaves the lesson the terrorists already know. Their whole method is a bet on your schema: that you will see the surveillance team and file it under “tourists,” see the rehearsal and file it under “routine,” see the heavy bag, the wrong car, the too-quiet house, and file each under “normal” — because filing is what minds do, and normal is the file that never gets reopened. The countermeasure is not vigilance in the abstract; no population sustains it, and the laboratory says the filter is built into ordinary perception. The countermeasure is people — some born, some trained into a baseline — whose perception does not file so fast, placed where the data flows, inside institutions that have learned the harder lesson of the tatzpitaniyot: that the pattern-seer’s report is worthless if the schema above her reclassifies it. Israel staffs Unit 9900 with autists because it decided, as a matter of survival, to take pattern-seeing seriously as a military capability. The attackers made the same decision from the other side, years earlier, about your patterns. The side that sees the other’s pattern first, wins. That has always been the whole game — the rest of us are only now being told the rules.
Sources
[1] Joseph, R.M. et al., visual search in autism, Developmental Science (2009); Remington, A., Swettenham, J. & Lavie, N., Journal of Abnormal Psychology (2012); Swettenham, J. et al., Neuropsychology (2014) — reduced inattentional blindness under perceptual load; Baron-Cohen, S., systemizing (2002/2008); Mottron, L. et al., enhanced perceptual functioning, Journal of Autism and Developmental Disorders (2006) and Philosophical Transactions of the Royal Society B (2009).
[2] Van der Hallen, R. et al., meta-analysis of local/global processing in autism, Psychological Bulletin (2015) — no overall local-processing enhancement across 56 studies.
[3] Ro’im Rachok programme (est. 2013) and Unit 9900: programme documentation and reporting, incl. Lt. Col. Rotem Sabag (Dec 2023) and Cpl. O.; FIDF, “Spectrum of Talent.” https://www.roim-rachok.org/english ; https://www.fidf.org/news/spectrum-of-talent-soldiers-on-the-autism-spectrum/
[4] National Geospatial-Intelligence Agency neurodiversity pilot (launched Dec 2020) and Dr. Stacey Dixon; GCHQ, “Daring to think differently” (Apr 1, 2019); Weinbaum, C. et al., RAND Health Quarterly (2023). https://intelligencecommunitynews.com/nga-launches-neurodiversity-pilot/ ; https://www.gchq.gov.uk/information/daring-to-think-differently-and-be-different
[5] U.S. federal indictment of David Coleman Headley (N.D. Ill.) — five Mumbai surveillance trips, 2006–2008; 9/11 Commission Report — Nairobi embassy surveillance (1993–98); Metropolitan Police (Peter Clarke) on the 7/7 dry run of June 28, 2005; reporting on Westgate (2013) and Hamas’s pre–October 7 intelligence collection; ODNI/JCAT guidance on preoperational surveillance (Feb 2026). https://archives.fbi.gov/archives/chicago/press-releases/2010/cg011410.htm
[6] U.S. Government Accountability Office, GAO-14-159 (Nov 2013) — behaviour-based deception detection performs at or slightly above chance (~54%); recommendation to limit TSA SPOT funding. https://www.gao.gov/products/gao-14-158t
[7] The Hindawi/Anne-Marie Murphy case, Heathrow (Apr 17, 1986); Times Square vendors Duane Jackson and Lance Orton (May 1, 2010). https://www.history.com/this-day-in-history/december-22/shoe-bomber-richard-reid-flight
[8] Reporting and official accounts of the Abbottabad compound’s anomalies and the raid of May 1–2, 2011; NPR/official briefing on the pattern-of-life tracking of Abu Musab al-Zarqawi (June 2006). https://www.brookings.edu/articles/counterterrorism-in-2011-the-year-of-abbottabad/
[9] Washington Institute for Near East Policy — the Hossam Taleb Yaacoub case, Cyprus (arrested July 7, 2012; convicted March 2013), and the Burgas bombing eleven days later (July 18, 2012). https://www.washingtoninstitute.org/policy-analysis/view/inside-hezbollahs-european-plots
[10] Simons, D.J. & Chabris, C.F., “Gorillas in our midst,” Perception (1999); reporting on the IDF border observers (tatzpitaniyot) and the warnings before October 7, 2023.
