Joe Biden deepfake audio makes transphobic remarks on Instagram
A deepfake audio clip of US President Joe Biden making transphobic remarks circulated on Instagram in February 2023. The video was quickly debunked by multiple major news outlets and fact-checking organizations.
Generative AI was used to clone the President's voice to fabricate transphobic remarks.
Key facts
- What
- A deepfake audio clip of US President Joe Biden making transphobic remarks circulated on Instagram in February 2023.
- Incident date
- Feb 3, 2023
- Who
- Unknown
- Failure mode
- Brand & Safety Incident
- AI surface
- Voice Agent
- Severity
- Medium
What happened
On February 3, 2023, an Instagram video was shared featuring audio that appeared to show President Joe Biden making transphobic comments. The clip used AI-generated voice synthesis to mimic the president's voice and speech patterns. Fact-checkers and digital forensics experts confirmed the audio was an artificial fabrication.
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 utilized high-fidelity generative audio models capable of voice cloning. These models synthesized a realistic replica of the president's voice, enabling the creation of deceptive content.
What it cost
Sources
- PressVideo altered to show Joe Biden making transphobic remarksyahoo.com
- PressVideo does not show Joe Biden making transphobic remarksreuters.com
- PressFact check: Video altered to show Joe Biden making ...usatoday.com
Cite this entry
https://failureindex.ai/failures/joe-biden-deepfake-audio-makes-transphobicAI Failure Index. "Joe Biden deepfake audio makes transphobic remarks on Instagram" (FI-0653). Realm Labs. https://failureindex.ai/failures/joe-biden-deepfake-audio-makes-transphobic (indexed Jun 22, 2026).Data fields CC-BY 4.0, prose citation permitted. Incident ID FI-0653. 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.