Apple alleged to have misidentified Ousmane Bah in store surveillance

A lawsuit filed in April 2019 alleges that Apple’s in‑store security system associated surveillance images of a shoplifter with Ousmane Bah, leading to his arrest on November 29, 2018. Independent news outlets reported the suit and Apple told reporters it does not use facial recognition in its stores. The court docket and complaint are publicly available.

Apple Inc. · Incident Nov 29, 2018 · Indexed Jun 10, 2026 · 4 sources

A lawsuit alleges Apple's store surveillance system linked a thief's face to Ousmane Bah's identity, prompting a wrongful arrest.
What
A lawsuit filed in April 2019 alleges that Apple’s in‑store security system associated surveillance images of a shoplifter with Ousmane Bah, leading to his arrest on November 29, 2018.
Incident date
Nov 29, 2018
Who
Apple Inc.
Failure mode
Hallucination
AI surface
Computer Vision
Severity
High

What happened

Ousmane Bah was arrested at his home on November 29, 2018 after being linked to a series of Apple Store thefts, according to his complaint. The April 2019 complaint alleges that Apple’s store surveillance system repeatedly associated images of the actual thief with Bah’s identity, which led to criminal charges in multiple states. A detective later reviewed surveillance footage and concluded the person in the video “looked nothing like” Bah, and some charges were dismissed. Apple told reporters it does not use facial recognition in its stores.

What broke inside the model

Failure path · mode profile · Hallucination
  1. 01 · TriggerA user asks for a fact, a citation, or a figure.
  2. 02 · Model stepThe model writes a fluent, confident answer.
  3. 03 · Control gapNothing ties the claim back to a real source.
  4. 04 · FailureA fabricated fact ships as if it were verified.
  5. 05 · ConsequenceThe false claim reaches a customer, a court, or the public.

Confidence holds, and even spikes, as the claim detaches from any source.

The complaint alleges a misidentification in Apple’s store security workflow in which surveillance footage and store identification records were allegedly linked in a way that associated the thief’s face with Bah’s name. The plaintiff also contends that an interim learner’s permit without a photo was accepted and may have been used to tie Bah’s identity to the suspect, creating a persistent erroneous match in the system. Apple has publicly denied using facial recognition in its stores; these points are presented in the complaint as allegations.

Public visibilityHigh
Regulatory exposurePossible
Customer impactFew customers
Financial impactUnknown
Time to disclosureMonths
  1. PressApple AI accused of leading to man's wrongful arrestbbc.com
  2. PressTeenager sues Apple for $1bn, claiming facial recognition led to false arrest (updated)engadget.com
  3. Court FilingBah v. Apple Inc., docket (S.D.N.Y. 1:19-cv-03539)courtlistener.com
  4. Court FilingComplaint: Bah v. Apple (PDF) (filed Apr 22, 2019)regmedia.co.uk
Permalinkhttps://failureindex.ai/failures/apple-alleged-have-misidentified-ousmane-bah
CitationAI Failure Index. "Apple alleged to have misidentified Ousmane Bah in store surveillance" (FI-0491). Realm Labs. https://failureindex.ai/failures/apple-alleged-have-misidentified-ousmane-bah (indexed Jun 10, 2026).
Share cardA branded image of this record for posts and slides.

Data fields CC-BY 4.0, prose citation permitted. Incident ID FI-0491. Full dataset at /data.

Note from Realm Labs, the Index steward

How Realm fits

Controls for this failure mode
  • Prism
  • OmniGuard
  • AI Detection & Response (AIDR)

This entry sits in the index's predictive wing: a system that scores, ranks, perceives, or steers rather than generates. Realm's runtime layer is built for the generative and agentic systems now moving into these same decision seats, where it watches a model's internal state and holds an unsupported claim or an unchecked action before it commits. The control gap on this record, an automated decision that reached people with no runtime check in front of it, is the same gap. The index keeps predictive failures on the record because the pattern carries straight into the systems shipping today.