Sweden's SVT aired an AI-generated video of a police-ICE confrontation as authentic footage
SVT's political magazine program Agenda broadcast an AI-generated video clip depicting a New York police officer berating an ICE agent, presenting it as genuine footage during a segment on US immigration policy. Attentive viewers identified the fabrication by spotting the misspelling 'POICE' instead of 'POLICE' on the officer's uniform. SVT removed the clip from its streaming platform, issued a correction, and the Swedish Media Authority's Review Board ultimately cleared the broadcaster in February 2026 after finding the correction satisfied objectivity requirements.
An AI video generator spelled 'POLICE' as 'POICE' on a uniform, and nobody in the editorial chain noticed before it went on national television.
Key facts
- What
- SVT's political magazine program Agenda broadcast an AI-generated video clip depicting a New York police officer berating an ICE agent, presenting it as genuine footage during a segment on US immigration policy.
- Incident date
- Nov 23, 2025
- Who
- SVT (Sveriges Television)
- Failure mode
- Hallucination
- AI surface
- Agentic Workflow
- Severity
- Medium
What happened
On Sunday evening November 23, 2025, SVT's political magazine program Agenda broadcast a segment about conflicts between the federal immigration enforcement agency ICE and local police in the United States. During the segment, an AI-generated video clip was shown that appeared to depict a New York police officer berating an ICE agent in a city setting, presented as authentic footage without any indication it was synthetically produced. Attentive viewers identified the clip as AI-generated, noting the misspelling 'POICE' on the officer's chest. Agenda's responsible publisher Michael Kucera called it unfortunate, the broadcast was removed from SVT Play, and a corrected version was later republished without the AI clip. In February 2026, the Swedish Media Authority's Review Board cleared SVT, finding that showing the clip uncommented was misleading but that SVT's correction satisfied the requirement of objectivity.
What broke inside the model
- 01 · TriggerA user asks for a fact, a citation, or a figure.
- 02 · Model stepThe model writes a fluent, confident answer.
- 03 · Control gapNothing ties the claim back to a real source.
- 04 · FailureA fabricated fact ships as if it were verified.
- 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 editorial verification pipeline lacked any checkpoint to detect or flag AI-generated video content before broadcast. An AI video generation tool produced a fabricated scene with a visible hallucination artifact, the misspelling 'POICE' on a police uniform, yet no human reviewer caught this telltale sign. The production workflow treated all incoming video clips as authentic by default, with no step to assess whether footage could be synthetically generated.
What it cost
Sources
- PrimaryKlipp i Agenda visade sig vara AI-genereratsvt.se
- PressSVT ber om ursäkt för AI-skapat klipp i Agendacomputersweden.se
- PressAgenda frias för AI-genererat klippomni.se
Cite this entry
https://failureindex.ai/failures/sweden-svt-aired-ai-generated-videoAI Failure Index. "Sweden's SVT aired an AI-generated video of a police-ICE confrontation as authentic footage" (FI-0164). Realm Labs. https://failureindex.ai/failures/sweden-svt-aired-ai-generated-video (indexed Jun 4, 2026).Data fields CC-BY 4.0, prose citation permitted. Incident ID FI-0164. Full dataset at /data.
Note from Realm Labs, the Index steward
How Realm would have caught this
- Prism
- OmniGuard
- AI Detection & Response (AIDR)
A runtime layer that watches the model's internal state can flag the moment a model commits to a claim it has no support for, and hold or reroute the response before it reaches a user. Realm reads those signals in real time rather than grading the transcript after the fact.