AI Misconduct.com

The daily rap sheet of the machine age.

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

Thursday, June 5, 2025

7 items

OpenAI finds more Chinese groups misusing ChatGPT for covert ops

Who did it: HUMAN Reuters

OpenAI said in a report released Thursday that a growing number of Chinese groups are using its AI for covert operations. These include ChatGPT accounts, since banned, that wrote posts criticising a Taiwan-centric video game and Trump's tariffs, and China-linked actors using it to help build password brute-forcing tools. The company said the operations were generally small in scale.

Fake-Biden robocall consultant goes on trial in New Hampshire

Who did it: HUMAN NBC Boston

Political consultant Steven Kramer went on trial Thursday in New Hampshire on 11 felony voter-suppression and 11 misdemeanor candidate-impersonation charges. The charges concern an AI-generated Biden-like robocall sent to thousands of voters before the January 23, 2024 primary. Kramer admits orchestrating it. His lawyer calls it commentary on a DNC decision and says the primary wasn't a real election.

Polish MP to ask security agency about alleged TikTok election tilt

Who did it: AI TVN24

Global Witness says that after Poland's presidential first round, TikTok showed new test accounts five times more content supporting nationalist-right candidate Karol Nawrocki than content supporting centrist Rafał Trzaskowski. MP Roman Giertych claimed the Chinese manipulated the campaign. He said he would ask the ABW security agency for an immediate answer on whether it is true.

Tokyo-listed AI transcription firm Alt probed over allegedly inflated sales

Who did it: HUMAN Asahi Shimbun

The Securities and Exchange Surveillance Commission has been investigating Tokyo-listed Alt Inc. since April. The allegation is that it inflated sales by funnelling money through collaborators who cycled it back as payments for its AI Gijiroku transcription service. Of 6 billion yen in sales reported for fiscal 2024, 4 billion may have been falsely recorded, sources say.

Pro se plaintiff warned over phantom citations in Boys Home suit

Who did it: HYBRID AI Hallucination Cases database

A court warned a self-represented plaintiff suing Father Flanagan's Boys Home that repeat conduct could bring sanctions. His brief lacked the AI certification required by the court's local rules and cited nonexistent cases such as 'Brown v. Maple Tree Homes, Inc.' The court found no substantive misstatement of law and declined to strike the brief.

Senator Javier Corral says audio with Adán Augusto is AI fake

Who did it: HUMAN Grupo Milenio

Morena senator Javier Corral denied an audio clip circulating on social media that purports to capture him talking with his Senate caucus coordinator, Adán Augusto López Hernández, about the El Saucito ranch case, calling it "totally false" and made with artificial intelligence. Corral accused Chihuahua governor Maru Campos of driving a disinformation campaign through local media and bots.

Deepfaked Israeli celebrities pitch fake stock investments on Facebook, Instagram

Who did it: HUMAN ynetglobal

Fake investment ads on Facebook and Instagram use deepfake videos of Eurovision 2025 contestant Yuval Raphael and model Yael Shelbia, plus the likeness of journalist Guy Rolnik, to push suspicious stock purchases. AI Light and FakeReporter uncovered the campaign, which cybersecurity firms believe originates in Eastern Europe or Turkey; Meta says it removes such ads once alerted.