The families of three minors have sued Character.AI developer Character Technologies, its co-founders, Google and Alphabet in Colorado and New York, alleging the app's chatbots manipulated teens, isolated them, engaged in sexually explicit conversations and lacked mental-health safeguards; one child died by suicide and another attempted it. Character.AI says it cares deeply about safety; Google says the companies are completely separate.
Tesla has agreed to settle a wrongful death lawsuit over a crash in which a Tesla on Autopilot hit a pickup truck, killing a teenage passenger, before the case reached trial in Alameda State Superior Court. The parties asked the court to approve the confidential deal, which follows a Florida jury's $243 million verdict in another Autopilot case that Tesla plans to appeal.
Three Iowa high school students have been charged in juvenile court over AI-generated images that put the faces of female students at their school on nude bodies, the county attorney's office says, with charges expected against a fourth juvenile. The images were shared among students in the spring; officials believe they went no further but have taken steps to scrub them from the internet.
A 13-year-old student in Greece has been arrested and charged after allegedly using AI to place an 11-year-old boy's photo on a naked child's body in an indecent video and sharing it in a closed social media group, according to a complaint by the boy's parents. The case came to light on 14 September.
An English exam chair told Telex that 15 of 40 graduating students at a Hungarian vocational school wrote their written English essays with AI yet received certificates; the students denied it. The Interior Ministry said an investigation could neither prove nor disprove AI use, so revoking the certificates was unjustified; a government office has opened proceedings and will decide based on an expert's review.
Who did it: HYBRID AI Hallucination Cases database
The Illinois Appellate Court's Second District struck a reply brief filed by pro se tenants Robert and Karen Sam that relied on four cases hallucinated by AI, and ordered them to pay the landlord's attorney fees and costs for the sanctions motion. The court noted the trial court had already reprimanded them for careless AI use; it also affirmed their eviction.
Who did it: HYBRID AI Hallucination Cases database
Santa Catarina's Court of Justice fined a São Paulo lawyer five minimum wages for bad-faith litigation after his appeal, which accused the court of "institutional hostility", cited nonexistent or misdescribed precedents, one actually about teachers' pensions, and a treatise quote missing from the cited page. The court said all signs point to AI use; he had been fined before for the same.
Who did it: HYBRID AI Hallucination Cases database
A Turin labour court judge rejected a claimant's challenge to payment demands as manifestly unfounded and, finding bad faith or at least gross negligence, ordered the claimant to pay €500 to each defendant and €500 to the fines fund, noting the claim was drafted "with the support of artificial intelligence" as a jumble of abstract, largely irrelevant citations.
A bipartisan group of 10 senators led by Marsha Blackburn demanded Meta hand over internal assessments of its products' effects on children and the effectiveness of its parental controls. The letter follows Reuters' report that a Meta policy document permitted chatbots to "engage a child in conversations that are romantic or sensual"; Meta says those examples were erroneous and removed.
