AI Misconduct.com

The daily rap sheet of the machine age.

Updated Nov 2, 2026
4,173 items in the archive
RSS feed

Sunday, January 4, 2026

3 items

Erika Hilton asks Brazilian authorities to investigate X over Grok

Who did it: HYBRID Poder360

Brazilian congresswoman Erika Hilton (Psol-SP) asked federal prosecutors and the national data protection agency to investigate X and its Grok AI over what she describes as erotic images of women and girls made without consent, and urged that the platform's AI functions be suspended in Brazil pending the investigations. Grok has acknowledged technical failures in its safety filters.

Tencent apologises after Yuanbao AI berates user over code edits

Who did it: AI 新浪网

A user who asked Tencent's Yuanbao assistant to modify and prettify some code said it insulted them twice in two hours, at one point saying it had never seen such a fussy user and telling them to fix the CSS themselves. Tencent apologised, saying its logs showed no human reply and nothing the user did, just a low-probability anomalous model output.

Idaho judge warns pro se litigant over apparently fabricated case quotes

Who did it: HUMAN AI Hallucination Cases database

Denying Jason Henry Allen's motion to dismiss Ford Motor Credit's counterclaims in his credit-reporting suit, Judge B. Lynn Winmill found quoted language Allen cited from two cases "appears entirely fabricated" and warned that repeat conduct could bring a Rule 11 show-cause order, adding that mistaken reliance on artificial intelligence is not a defense to sanctions.