An anti-deepfake declaration may have been written by AI

A national suit implicit Minnesota’s “Use of Deep Fake Technology to Influence An Election” instrumentality is present straight dealing with the power of AI. In a caller filing, attorneys challenging the instrumentality accidental an affidavit submitted to enactment it shows signs of containing AI-generated text. The Minnesota Reformer reports Attorney General Keith Ellison asked Stanford Social Media Lab founding manager Jeff Hancock to marque the submission, but the papers filed includes non-existent sources that look to person been hallucinated by ChatGPT oregon different ample connection exemplary (LLM).

Hancock’s affidavit cites a 2023 survey published successful the Journal of Information Technology & Politics titled “The Influence of Deepfake Videos connected Political Attitudes and Behavior.”

But according to the Reformer, determination is nary grounds of that survey successful the Journal of Information Technology & Politics oregon immoderate different publication. Another root cited successful Hancock’s declaration, “Deepfakes and the Illusion of Authenticity: Cognitive Processes Behind Misinformation Acceptance,” doesn’t look to beryllium either.

Hancock did not respond to The Verge’s petition for comment.

“The citation bears the hallmarks of being an artificial quality (AI) ‘hallucination,’ suggesting that astatine slightest the citation was generated by a ample connection exemplary similar ChatGPT,” lawyers for Minnesota authorities Rep. Mary Franson and Christopher Khols — a blimpish YouTuber who goes by Mr Reagan — wrote successful a filing. “Plaintiffs bash not cognize however this hallucination coiled up successful Hancock’s declaration, but it calls the full papers into question, particularly erstwhile overmuch of the commentary contains nary methodology oregon analytic logic whatsoever.”