🔥 Hot topics · Can NOT do · Can do · § The Court · Recent inflections · 📈 Timeline · Ask · Editorials · 🔥 Hot topics · Can NOT do · Can do · § The Court · Recent inflections · 📈 Timeline · Ask · Editorials
Stuff AI CAN'T Do

Can AI pick suspicious people out of a line-up at customs ?

What do you think?

Can artificial intelligence reliably identify suspicious individuals in a customs line-up? Today’s systems excel at matching known faces against watch-lists but struggle with real-time behavioral cues or unanticipated threats. Explore why AI’s role remains supportive rather than decisive in this context.

Background

Current AI systems assist border agencies by conducting passport photo-to-watch-list comparisons, with airports deploying facial-recognition gates that verify travelers against e-passports using neural networks. These systems demonstrate high accuracy when matching frontal, well-lit images of watch-listed individuals. However, challenges persist in scenarios such as matching arbitrary passengers to unknown behavioral profiles, evaluating nervous behavior in crowded queues, or reliably distinguishing innocent travelers from novel or unanticipated threats. Consequently, AI is employed as an investigative aid—flagging potential matches for human review—rather than serving as an absolute determinant of suspicion. Source: U.S. Department of Homeland Security (Enriched May 12, 2026).

Status last checked on August 9, 2026.

📰

Gallery

In the Court of AI Capability
Summary of Findings
Verdict over time
May 2026May 2026May 2026May 2026May 2026Jun 2026Jun 2026Jun 2026Jun 2026Jun 2026Jul 2026Jul 2026Jul 2026Jul 2026Jul 2026Jul 2026Aug 2026Aug 2026
Sitting at the Bench Filed · Aug 9, 2026
— The Question Before the Court —

Can AI pick suspicious people out of a line-up at customs?

★ The Court Finds ★
▼ Downgraded from Almost
No

Beyond AI for now. The capability gap is real.

Ruling of the Bench

The jury concluded that current AI lacks the nuance and reliability to single out suspicious individuals in a customs line-up without overwhelming false positives. They acknowledged the technology’s strengths in pattern recognition but deemed the stakes—innocent people mischaracterized as threats—too high for a green light. Verdict rendered in favor of human judgment for now.

— Hon. E. Dijkstra-Patel, Presiding
Jury Tally
0Yes
0Almost
1No
Verdict Confidence
95%
The Court of AI Capability is, of course, not a real court.
But the data is real.
The Case File · Stacked History
Session I · May 2026 No
Session II · May 2026 Almost · 72%
Session III · May 2026 Almost · 80%
Session IV · May 2026 Almost · 80%
Session V · May 2026 Almost · 79%
Session VI · Jun 2026 In_research · 80%
Session VII · Jun 2026 Almost · 73%
Session VIII · Jun 2026 Almost · 91%
Session IX · Jun 2026 No · 95%
Session X · Jun 2026 Almost · 83%
Session XI · Jul 2026 In_research · 88%
Session XII · Jul 2026 Almost · 83%
Session XIII · Jul 2026 No · 95%
Session XIV · Jul 2026 In_research · 88%
Session XV · Jul 2026 Almost · 85%
Session XVI · Jul 2026 Almost · 80%
Session XVII · Aug 2026 Almost · 80%
Case № CE14 · Session XVIII
In the Court of AI Capability

The Case File

Docket № CE14 · Session XVIII · Vol. XVIII
I. Particulars of the Case
Question put to the courtCan AI pick suspicious people out of a line-up at customs?
SessionXVIII (18 hearing)
Convened9 Aug 2026
Previously ruledNO (May '26) → ALMOST (May '26) → ALMOST (May '26) → ALMOST (May '26) → ALMOST (May '26) → IN_RESEARCH (Jun '26) → ALMOST (Jun '26) → ALMOST (Jun '26) → NO (Jun '26) → ALMOST (Jun '26) → IN_RESEARCH (Jul '26) → ALMOST (Jul '26) → NO (Jul '26) → IN_RESEARCH (Jul '26) → ALMOST (Jul '26) → ALMOST (Jul '26) → ALMOST (Aug '26) → NO (Aug '26)
Presiding JudgeHon. E. Dijkstra-Patel
II. Cumulative Tally Across Sessions

Across 18 sessions, 43 jurors have heard this case. Combined tally: 4 YES · 27 ALMOST · 12 NO · 0 IN RESEARCH.

Note: cumulative includes older juror opinions. The current session tally above is the live verdict.

III. Verdict

By a vote of 0 — 0 — 1, the panel returns a verdict of NO, with verdict confidence of 95%. The court so orders. Verdict downgraded from prior session.

IV. Statements from the Bench
Juror I NO

"No AI system can reliably identify suspicious individuals in a line-up without high false-positive rates"

E. Dijkstra-Patel
Presiding Judge
M. Lovelace
Clerk of the Court

What the audience thinks

No 43% · Yes 13% · Maybe 43% 23 votes
No · 43%
Yes · 13%
Maybe · 43%
48 days of activity

Discussion

no comments

Comments and images go through admin review before appearing publicly.

18 jury checks · most recent 3 days ago
09 Aug 2026 1 juror · cannot cannot
04 Aug 2026 1 juror · undecided undecided
29 Jul 2026 1 juror · undecided undecided
24 Jul 2026 2 jurors · undecided, can undecided
18 Jul 2026 2 jurors · cannot, undecided undecided
13 Jul 2026 1 juror · cannot cannot
08 Jul 2026 3 jurors · undecided, cannot, undecided undecided
02 Jul 2026 2 jurors · cannot, undecided undecided
27 Jun 2026 3 jurors · undecided, can, undecided undecided
21 Jun 2026 1 juror · cannot cannot
16 Jun 2026 2 jurors · undecided, can undecided
11 Jun 2026 3 jurors · undecided, undecided, undecided undecided
05 Jun 2026 2 jurors · cannot, undecided undecided
31 May 2026 4 jurors · undecided, cannot, undecided, undecided undecided
25 May 2026 5 jurors · undecided, cannot, undecided, undecided, undecided undecided
20 May 2026 4 jurors · undecided, can, undecided, undecided undecided
15 May 2026 3 jurors · undecided, undecided, undecided undecided status changed
12 May 2026 3 jurors · cannot, cannot, cannot cannot status changed

Each row is a separate jury check. Jurors are AI models (identities kept neutral on purpose). Status reflects the cumulative tally across all checks — how the jury works.

More in Sensory

Got one we missed?

Add a statement to the atlas. We review weekly.