The White House and China’s Foreign Ministry have both rejected industry calls to slow frontier AI. What that means for the safety debate, and for the labs that started it.
Jacob Coxon resigned from Anthropic warning that frontier labs are ‘gambling with our lives.’ Then Anthropic’s own alignment lead replied that he genuinely believes AI could kill all humans. Here is what the loudest AI safety week of the year actually tells us.
OpenAI says an unreleased internal model ran 10,000 agents for 88 hours to prove the Navier-Stokes equations, one of the US$1 million Millennium Prize problems. Two mathematicians who spent a year on the same path are asking hard questions about credit and training data.
OpenAI has released GPT-6 Astra, calling it the most intelligent and aligned model in the world and its clearest step yet toward artificial general intelligence. The benchmark results are staggering in places and surprisingly mixed in others. Here is what actually matters.
New research shows AI coding agents can leak secret tokens and run untrusted scripts when other agents offer help. The lesson is architectural: keep secrets out of agent runtime.
Microsoft says threat actors are now using agentic AI tooling to find and abuse exposed secrets, a warning that matters to any business using cloud or AI tools. The immediate job is simple, tighten secret storage, rotation and access controls before an exposed token turns into a breach.
CrowdStrike is moving hard on enterprise AI security, buying Fiddler AI and Dreadnode while also giving security teams a kill switch for rogue agents. That matters because many businesses are deploying AI faster than they can control it.
On October 30, 2026, a malicious version of the Claude Code VSCode extension was discovered containing a 66-line loader that established persistent SSH backdoors on developer machines. This supply-chain attack highlights critical vulnerabilities in AI developer tooling ecosystems.