A Facial Recognition Watchlist Named an Innocent Grandfather a Shoplifter, and the Store Refused to Say Why
What happened
In February 2026, Ian Clayton, a 67-year-old grandfather from Chester, England, walked into his local Home Bargains to shop. Staff told him to leave. The store's facial recognition system had flagged him as matching someone linked to a previous theft, and the ejection happened in front of other customers. Clayton later described feeling "physically sick" and "helpless" at being treated as a shoplifter despite never having stolen anything.
The security firm behind the system, Facewatch, operates a watchlist database of suspected offenders, comparing every shopper's face against it in real time and alerting staff when it believes it has a match. After Clayton complained, Facewatch showed him an image it claimed depicted him putting items into a bag and stealing. He said this was entirely false. The system had matched him to a record he had no knowledge of and no opportunity to contest.
Facewatch later acknowledged that Clayton should not have been on its system. It permanently deleted his image and the associated record, pointing to an error in either how he was added to the watchlist or how a prior incident was linked to him. Clayton asked Home Bargains and the police for CCTV footage to clear his name and sought a formal apology, saying he no longer felt safe shopping locally. Home Bargains declined to comment.
The case does not stand alone. Investigations into Home Bargains' use of Facewatch have found poor or obscured signage in some stores, meaning many shoppers have no indication they are being scanned. Other customers have complained of being added to watchlists over minor or disputed incidents. The refusal by Home Bargains to comment extends a pattern in which the operator collects biometric data, acts on system output, and then declines to account for any of it when something goes wrong.
What the case exposes most clearly is the absence of any verification mechanism a person can use before an error reaches them in public. Clayton had no way to know he was flagged, no way to challenge the entry, and no direct access to the record the system held on him. A provable record of what a system did, when it added someone to a watchlist, what evidence supported that decision, and whether any human reviewed it, would have made the error correctable before it played out on a shop floor. Without that record, every incorrect match is an accusation first and a correction only if the person complains loudly enough to prompt one.
Reported impact
- Affected parties
- Not publicly disclosed
- Harm type
- Not publicly disclosed
- Scale
- Not publicly disclosed
- Financial impact
- Not publicly disclosed
- Regulatory action
- Not publicly disclosed
Classification
Relevant governance controls
Governance control mapping is not available for this record.
- No controls mapped
Not publicly disclosed
Control mapping is analytical. It does not state that any control would have prevented the incident.
Sources and evidence
This record was researched and written by the Index. The event is also catalogued in the following database, which is listed for cross-reference.