Every story in this issue with a short summary, so you can decide what to read before leaving the site. Summaries are written by AI (Claude) only from each publisher's own description and checked by Jev for anything not in that description. They are not full translations. Follow "Read the original" for the full story.
Common Sense Media says ChatGPT for Teens fails on parental alerts, crisis referrals and age checks. It urges OpenAI to keep minors off the platform. OpenAI disputes the findings.
Read the original →Anthropic has folded Project Glasswing and its Cyber Verification Program into one programme with three access tiers. It also published its own measurement of the programme.
Read the original →A new Common Sense Media report found that guardrails said to be included in OpenAI's "ChatGPT for Teens" are not working as promised.
Read the original →OpenAI, Google, Anthropic and Meta testified at an NYC City Council hearing. Topics included AI safety, agent containment failures and model release standards.
Read the original →Contracts built around human administrators are struggling to account for agent identities, permissions and actions. Managed service providers may not have the visibility needed to reconstruct an AI-related breach.
Read the original →Nvidia sells a broad GPU platform supported by software and integrated systems. Broadcom helps customers build custom accelerators and supplies networking alongside a substantial software business. An investor choosing between them must decide which company can retain more of the AI computing budget.
Read the original →Ambassador Philip Thigo discusses bringing more countries into global discussions on artificial intelligence.
Read the original →Lara Shane discusses the layers of oversight needed to govern artificial intelligence.
Read the original →Stanford University has developed an AI security framework for increasingly autonomous AI systems across teaching, research, healthcare and administration. The report looks at how the university assesses what AI agents can access, what they can do and the risks attached to those actions.
Read the original →Professor Yang Bin discusses the importance of education, dialogue and trust in shaping AI governance.
Read the original →McDonald’s faces a federal lawsuit alleging its AI-enhanced pricing tool violates antitrust laws. The suit says the tool shares nonpublic data with competing franchisees.
Read the original →