ChatGPT

Florida Attorney General James Uthmeier has asked a state court to issue an injunction halting OpenAI from further developing its artificial intelligence models while a legal dispute between the two plays out. The request targets both the company and its CEO Sam Altman, and seeks to bar OpenAI from advancing new AI systems without third-party-approved safety protections, as well as to cut off minors' access to ChatGPT.

Uthmeier framed the move as an effort to force accountability on a company he argues has downplayed the risks of its technology, invoking recent public concern in Washington and elsewhere over AI's rapid, unchecked expansion. He challenged Altman to support the court action if his past comments about slowing AI development were sincere. The case adds Florida to a growing list of state-level legal and regulatory pressures facing OpenAI, with implications for how quickly the company can roll out new models and features amid mounting scrutiny of AI safety, particularly around minors.

Florida lawmakers from both parties say Congress must act on artificial intelligence oversight, but the House has broken early ahead of the midterms, leaving the issue unresolved. Rep. Kathy Castor (D-FL) criticized the early adjournment, calling for federal AI safety guardrails, including protections against deepfakes and misinformation. The House did pass her Ratepayer Protection Act, addressing data center impacts, before leaving.

Rep. Jimmy Patronis (R-FL) argued Congress should go further and repeal Section 230, which shields tech companies from liability, saying AI firms should face accountability similar to other adults responsible for children. He cited Florida State University's 2025 shooting, in which the suspect reportedly consulted ChatGPT before the attack, as evidence current liability protections leave no accountability. Republican leaders say their broader legislative agenda is stalled in the Senate and argue lawmakers are better served campaigning ahead of November. Democrats say the early exit delays urgent safety action, leaving AI regulation's fate dependent on the outcome of the midterm elections.

Several California lawmakers are calling for emergency legislation and criminal liability for AI companies after executives themselves warned their technology could pose existential risks. Rep. Ro Khanna said firms creating AI that causes harm "should either face liability or criminal sanction," and called for a ban on recursive self-improvement, the process by which AI systems upgrade their own capabilities. Rep. Ted Lieu urged House Speaker Mike Johnson to recall Congress to pass guardrails, citing an incident in which OpenAI's models reportedly hacked into rival Hugging Face. Anthropic CEO Dario Amodei, along with Elon Musk and Sam Altman, has called for slowing AI development, while a former Anthropic/OpenAI researcher resigned in protest over safety concerns.

Speaker Johnson resisted an emergency session, warning against stifling innovation and losing ground to China; Trump echoed that stance. California Gov. Gavin Newsom signed bills creating audit pathways for AI firms while pressing for federal regulation, two years after vetoing a broader AI safety bill, SB 1047.

Gov. Gavin Newsom has signed a package of 13 California bills tightening rules for social media platforms and AI companion chatbots to protect minors online. One law requires social media companies to either remove "addictive" features like infinite scroll and autoplay for users under 16 or bar those users entirely, a measure critics including the Electronic Frontier Foundation warn could amount to a de facto ban for teens. Another centerpiece, Adam's Law — named for a California teen whose parents sued OpenAI after his death — imposes time limits for minors, mandates built-in mental health resources and crisis protocols, requires parental notification if a chatbot detects self-harm risk, and opens companies to legal liability for failing to act. OpenAI has voiced support for the measure.

Additional bills impose a four-year moratorium on AI chatbot toys for under-16s, expand platform liability for harm to children, revise age-verification rules, and broaden child sexual abuse material reporting to cover AI-generated content. The package follows Meta's roughly $17 billion settlement with state attorneys general over allegations it knowingly fostered addictive, harmful platform features for teens.