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.
The Federal Trade Commission has asked executives from Anthropic to testify as part of its consumer harm probe into the maker of the Claude AI model. An FTC spokesperson confirmed the request.
Read the original →The report found that legislators in 30 states enacted 115 AI-related laws this year, as Congress declined to act. The laws include chatbot safety laws, frontier AI model protocols, and new definitions of personhood.
Read the original →Researchers propose a scientist-led international network to advance AI safety science, tools and standards. They argue that current multilateral efforts are too slow and constrained.
Read the original →A founder detailed in a recent social media post how an AI agent destroyed much of his company's production data.
Read the original →Agent decision logs and telemetry data belonging to users of Warp's agentic development environment were exposed. The exposure came from an attack on a third-party vendor in the project's infrastructure.
Read the original →Chargebacks911 has launched indemnification for merchants against qualifying chargebacks on purchases made by AI agents.
Read the original →A configuration mismatch in GLM 5.3 on Workers AI triggered an SGLang bug that exposed KV cache pages across requests. The post describes how the issue was addressed.
Read the original →TSMC is reported to be considering a 20% price increase for CoWoS packaging. AI demand is straining packaging capacity.
Read the original →Blockmate Ventures has entered a binding agreement with Scannit Inc. to acquire its first NVIDIA B200 AI compute node, under an agreement dated October 27, 2026. The node has eight NVIDIA B200 GPUs and costs approximately US$635,000. The turnkey price includes shipping, data centre setup, installation, network connectivity and onboarding.
Read the original →In a recent SEC filing, Anthropic highlighted the risks of relying on TSMC for AI chip production. It warned that disruptions could severely affect its business and the broader AI industry.
Read the original →Mitsubishi Electric announced it obtained ISO/IEC 42001 certification from third-party certification body SGS Japan Inc. The certification covers its AI Agent Platform, which supports AI use across the Mitsubishi Electric Group. The standard specifies management requirements for organizations to manage AI-specific risks.
Read the original →As organizations integrate AI into their business operations, they need clear ways to manage the risks associated with its development and use. Effective governance provides oversight.
Read the original →Testing by BixeLab found that ADVANCE.AI's liveness check stopped all 125 spoofing attempts using photos, masks, and video.
Read the original →An appellate court threw out the prison sentence of a man convicted of manslaughter. The ruling concerned the use of an AI-generated video depicting his slain victim.
Read the original →Von der Leyen cited escaping AI agents and the Hugging Face incident, then invited frontier labs to talk. She set no date.
Read the original →