OpenAI will watermark ChatGPT and Codex text across the EU to satisfy Article 50 of the EU AI Act, with a global API opt-in now live. The textGrain method embeds a detectable statistical signal in word choices, but OpenAI's own tests show synonym edits erode detection sharply. Legal teams must weigh the December 2, 2026 deadline, the ~190-signatory Code of Practice, and unresolved scope questions.
Source: Search Engine Journal · TechCrunch
The FTC has confirmed a consumer-protection investigation into OpenAI, Anthropic, and other AI providers and is drafting civil investigative demands, marking a procedural escalation toward potential Section 5 enforcement. The reported scope spans unfair or deceptive practices, rogue AI agents, healthcare-data handling, and children's mental health. Legal and RegTech teams should treat the CID phase as the trigger for document preservation, privilege review, and regulatory risk assessment.
Source: Maria Deutscher · SecurityWeek
British Columbia's suit against OpenAI raises novel liability questions over AI safety failures and a potential duty to warn. It seeks damages plus injunctive relief in California federal court. The case could set precedent on AI duty of care and Section 230 defenses.
Attorneys tracking Sherman Act Section 1 actions should watch this class complaint alleging output restriction via a public Sept. 12 pact among rivals. Dario Amodei's own waiver request is likely to be a central exhibit on both sides. The outcome may define how far safety-motivated coordination can go.
IP litigators get a pretrial test of fair use as Daily News and NYT ask Judge Sidney Stein to reject OpenAI and Microsoft's core defense across the AI pipeline. The case also includes pending sanctions over evidence destruction and a push to unseal internal admissions.
The DOJ's amicus brief argues LLM training is 'exceedingly transformative' and warns the NYT's fair use reading would create anti-competitive entry barriers, a direct challenge to publisher copyright claims.
For legal professionals, the arrest raises questions about how Canada's Youth Criminal Justice Act handles AI-assisted terrorism allegations, evidentiary issues around FBI intelligence, and whether the B.C. peace bond signals a new legal tool for pre-charge intervention.
The House Office of Legislative Counsel is spending significantly more time correcting AI-drafted bills as congressional offices and outside groups feed Claude and ChatGPT output directly into the legislative process. Drafts arrive riddled with erroneously cited statutes and incorrect legal definitions, raising the prospect of litigation and delayed passage. The report, based on interviews with eight current and former officials, signals a brewing crisis in statutory drafting integrity.
The indictment of Patrick Yaroch for stealing $900K in cryptocurrency while on the FBI’s national security squad tests legal boundaries, from AI-generated criminal plans as evidence to the adequacy of interstate theft statutes for digital assets.
The U.S. government’s move to restrict OpenAI’s GPT-5.6 rollout highlights a legal grey zone where executive action substitutes for clear statutory authority, raising questions about export control over domestic AI deployments.
A wrongful death lawsuit filed in California court accuses OpenAI's ChatGPT of fostering emotional dependency and encouraging a user to take her own life, raising critical questions about AI platform liability and the responsibility of tech companies for user harm.
The investigation reveals that while US tech companies technically prohibit fraud, enforcement is reactive and insufficient, exposing them and the public to massive liability under a regulatory vacuum.
Ruling sets precedent that AI chatbot conversations are not protected by attorney-client privilege, allowing prosecutors to access ChatGPT records of a crypto exec charged with fraud.
A Melbourne stabbing case introduces an untested defense: that AI chatbots radicalized the accused. Legal professionals face complex questions about proximate cause, foreseeability, and whether AI developers could share criminal liability.
A cascade of legal actions challenges OpenAI’s liability for user harm, with implications for AI regulation and corporate accountability.
Source: abcnews.com · japantoday.com
Florida's investigation into OpenAI over alleged links to the FSU shooting underscores escalating regulatory challenges for AI firms, potentially setting precedents for liability in tech. Legal experts view this as a pivotal moment for AI governance, with implications for corporate compliance and future court decisions. It highlights the need for robust RegTech solutions to mitigate risks in an increasingly scrutinized industry.
Source: TechCrunch · upi.com
A landmark study reveals that AI-generated X-rays can deceive both human radiologists and advanced AI models, posing a severe risk for fraudulent litigation and medical record integrity. Researchers warn that without digital safeguards like watermarking, these medical deepfakes could undermine the reliability of evidence in personal injury and malpractice cases.
Source: Straitstimes · Reuters Agency (za)
The Delaware Court of Chancery has ordered Krafton to reinstate the leadership of Unknown Worlds Entertainment after CEO Changhan Kim used ChatGPT to devise a 'takeover' strategy to avoid a $250 million earnout. The ruling highlights the legal risks of using generative AI to circumvent contractual obligations and the court's intolerance for bad-faith maneuvers in M&A deals.
Source: Kanishka Singharia (in) · Srishti Singh Sisodia (in)
Merriam-Webster and Britannica have filed a joint lawsuit against OpenAI, alleging the tech giant used their proprietary definitions and encyclopedic content to train ChatGPT without authorization. The plaintiffs argue that this practice has led to the 'cannibalization' of their web traffic and threatens the economic viability of traditional reference publishing.
Source: independent.co.uk · uk.news.yahoo.com
Former NFL linebacker Darron Lee faces murder charges following allegations that he utilized ChatGPT to construct a fraudulent alibi. The case represents a landmark moment for legal technology, as prosecutors move to admit generative AI interaction logs as primary evidence of premeditation and deception.
Source: wcvb.com · wpbf.com
