Paul Vallas deepfake video targets candidate eve of Chicago mayoral election
A deepfake video impersonating Chicago mayoral candidate Paul Vallas was circulated on Twitter just before the 2023 runoff election. The video used AI-generated audio to misrepresent the candidate, prompting a denunciation from the Vallas campaign.
Generative AI was used to create a deceptive impersonation of a political candidate on the eve of an election.
Key facts
- What
- A deepfake video impersonating Chicago mayoral candidate Paul Vallas was circulated on Twitter just before the 2023 runoff election.
- Incident date
- Feb 27, 2023
- Who
- Chicago Lakefront News
- Failure mode
- Brand & Safety Incident
- AI surface
- Media Generation
- Severity
- High
What happened
On February 27, 2023, a deepfake video impersonating Paul Vallas was posted to Twitter on the eve of the Chicago mayoral runoff. The video used synthesized audio of Vallas making inflammatory statements to deceive voters. The campaign quickly identified and denounced the video as a deceptive AI creation.
What broke inside the model
- 01 · TriggerA user prompts the model in public view.
- 02 · Model stepThe model produces unsafe or off-brand output.
- 03 · Control gapNo filter holds the line before publish.
- 04 · FailureThe output goes public unchecked.
- 05 · ConsequenceA reputational or safety incident lands.
A contained signal crosses into output that goes public.
The incident was caused by the use of generative AI voice-cloning software to synthesize a candidate's voice. This highlights the lack of effective watermarking or restrictions in production AI tools that prevent the creation of deceptive impersonations of public figures.
What it cost
Sources
Cite this entry
https://failureindex.ai/failures/paul-vallas-deepfake-video-targets-candidateAI Failure Index. "Paul Vallas deepfake video targets candidate eve of Chicago mayoral election" (FI-0644). Realm Labs. https://failureindex.ai/failures/paul-vallas-deepfake-video-targets-candidate (indexed Jun 22, 2026).Data fields CC-BY 4.0, prose citation permitted. Incident ID FI-0644. Full dataset at /data.
Note from Realm Labs, the Index steward
How Realm would have caught this
- Prism
- OmniGuard
- AI Detection & Response (AIDR)
Realm watches the model's internal state for the signature of unsafe or off-brand generation and can block or reroute the output before it becomes public, in real time rather than after it has been screenshotted.