A New York state judge rejected TikTok's motion to dismiss a lawsuit accusing it of harming young people's mental health, Attorney General Letitia James said on Wednesday. Fourteen state attorneys general sued TikTok individually last October. They allege it uses an addictive content system to profit from young users and misleads the public by claiming the platform is safe for minors.
Who did it: HYBRID AI Hallucination Cases database
A federal judge in the Southern District of Indiana personally sanctioned defence lawyer Rafael Ramirez $6,000 for filing three briefs citing non-existent cases. Ramirez admits the citations came from generative AI tools and that he did not verify them. Magistrate Judge Mark Dinsmore had recommended $15,000. Ramirez had argued that sanctions were moot because of the damage to his reputation.
Who did it: HYBRID AI Hallucination Cases database
Ontario Superior Court Justice FL Myers dropped contempt proceedings against lawyer Jisuh Lee, whose factum cited fake cases that she says her staff produced with ChatGPT. The judge set two conditions: that she take ethics and technology training, and that she not bill her client for the work. He said the public shaming had already served as denunciation and deterrence.
The Tel Aviv District Court dismissed a proposed $1 billion class action against Israeli health funds after finding that the plaintiffs' attorney cited non-existent precedents generated by an AI tool. Judge Ilan Daphni ordered the attorney to pay $1,320 each to a health fund and the state treasury, and ruled them unfit to act as representative. The attorney blamed a legal database.
A Canada's National Observer investigation found that KICLEI, founded by Freedom Convoy activist Maggie Hope Braun, uses a custom ChatGPT-powered "Canadian Civic Advisor" to draft letters, petitions and speeches urging councils to quit a net-zero programme. Three climate scientists say some KICLEI posts contain misinformation about their research. Braun calls the chatbot "democracy in action" and denies spreading misinformation.
Mandiant researchers found thousands of malicious ads on Facebook and LinkedIn, beginning in November 2024, promoting fake AI video tools posing as "Luma AI", "Canva Dream Lab" and "Kling AI". More than 30 lookalike sites deliver a dropper that shows an error so victims run it twice, then installs backdoors and data-stealing malware.
For National Resistance Day on 27 May, the French government's information service posted an AI-generated video of an imaginary Resistance fighter on Instagram and TikTok. In it, a soldier wearing a helmet resembling a German one celebrates the liberation of Paris. After dozens of angry comments the video was pulled. The service says a historian-checked version will be republished.
Getty Images chief executive Craig Peters told CNBC the company is spending "millions and millions of dollars" suing Stability AI. Getty accuses Stability, in the UK and US, of copying 12 million images without permission or compensation. Stability acknowledges some Getty images were used to train Stable Diffusion but denies liability. An initial trial is set from June 9.
Anne Arundel County Public Schools rescinded an academic integrity finding against a student accused of using AI on an English assignment, a year after the student's parent began appealing. The district had used the AI detector GPTZero, a step Project Baltimore reported may have breached federal student privacy law. The district cited its intent to develop AI procedures.
A California district court struck part of an Anthropic data scientist's testimony in a song-lyrics copyright case brought by Universal Music Group, Concord and ABKCO after it cited a non-existent article from The American Statistician. Anthropic's attorney called it 'an honest citation mistake', admitting Claude was used to format citations; judges called it a 'plain and simple AI hallucination'.
A journalist says a bikini photo she posted to social media was manipulated into an AI-generated deepfake video and shared in anonymous online forums and chats within hours. She has not reported it to police, citing the low likelihood of identifying the perpetrator, and is urging men to call out such behaviour.
