Full Summary
This Friday morning, OpenAI is delaying the public release of its latest AI model, GPT-5.6. Both Cyberpress.org and 41NBC News confirm that the Trump administration has requested OpenAI limit initial access to government-approved partners, citing national security concerns. OpenAI CEO Sam Altman informed employees that GPT-5.6 will first be available only to a limited group of enterprise partners, with the government approving access customer by customer. This decision follows a similar intervention with rival AI lab Anthropic, whose models were taken offline due to concerns over advanced cybersecurity capabilities. While OpenAI built in robust safety protections, including 700,000 GPU hours for automated red-teaming, this marks a significant shift in federal oversight over advanced AI releases in the United States. A broader public rollout is expected a couple of weeks after this limited preview. Meanwhile, a United States insurer is suing the creator of ChatGPT, alleging the chatbot engaged in the unauthorized practice of law. QLSPROCTOR.com.au reports that Nippon Life Insurance Company of America filed the suit after a litigant used ChatGPT to prepare at least 44 filings in a dispute against them. OpenAI publicly states the complaint lacks merit. In other developments, Viber is integrating ChatGPT directly into its messaging platform for users in the Philippines, allowing access to AI features like chat summaries and image editing without leaving the app. OpenAI has also updated GPT-5.5 Instant, focusing on making it a better conversational partner, and is preparing a new bidirectional voice model, Bidi 1, which allows ChatGPT to listen and speak simultaneously. However, the use of AI in personal matters carries risks. RTE.ie notes that millions are trusting AI chatbots with medical symptoms, with four out of five people considering using ChatGPT for self-diagnosis, despite its lack of awareness of potential mistakes. And in the legal field, Nerdbot reports lawyers are facing sanctions for using AI that "hallucinates" non-existent legal cases, underscoring the need for human verification. These developments mean that while AI tools are becoming more integrated into daily life, from messaging to legal work, users must navigate evolving government oversight and the inherent limitations of the technology.