AI Misconduct.com

The daily rap sheet of the machine age.

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

Sunday, June 1, 2025

4 items

Deepfake Senate president pitches fake Neptun Deep investments

Who did it: HUMAN HotNews.ro

Romania's National Cyber Security Directorate (DNSC) warned that a fake AI-made video shows Senate president Ilie Bolojan promoting "investments" in the Neptun Deep gas project, promising regular payments to anyone investing from 1,250 lei. According to DNSC, criminals alter existing public clips of politicians to lure victims into handing over data or money.

Town government's AI clone of scholar Wang Liqun sells fitness

Who did it: HUMAN 新浪网

A video account run by the government of Changgang town, in Henan's Sui county, posted an unauthorised AI-generated video of scholar Wang Liqun on May 31 plugging fitness methods and hailing a "once-in-60-years" auspicious Dragon Boat Festival. By June 1 the account was deleted, those responsible were summoned for talks and the town ended its deal with the operating company.

Israeli court fines plaintiff who blamed Google for phantom precedents

Who did it: HYBRID AI Hallucination Cases database

A small claims court in Hadera fined a plaintiff suing Hillel Yaffe Medical Center ₪250 on 1 June 2025 after Judge Yaniv Heller could not find two cited appeal rulings and doubted they existed. The plaintiff, who had vouched for ChatGPT's reliability, blamed the case-number errors on Google and an external search provider.

Council fed school phone survey comments into three chatbots

Who did it: HUMAN Glasgow Times

East Dunbartonshire Council fed open-text answers from its school smartphone surveys, which drew more than 11,000 responses including from young children, into ChatGPT, Gemini and Microsoft Copilot for summarising, without telling respondents. Formal complaints allege the surveys were flawed, and the council did not answer whether comments were checked for identifying details.