Tag
Discusses the evolving safety review process for frontier AI models, referencing Claude Fable 5's re-release and the need for a shared industry framework to assess jailbreaks, while expressing cautious optimism about the balance between safety and innovation.
Anthropic is redeploying Claude Fable 5 globally with new classifiers to block cybersecurity tasks, while collaborating with US government and industry partners on AI safety frameworks.
OpenAI announces collaborative security improvements with US CAISI and UK AISI, highlighting joint red-teaming efforts that discovered and helped remediate novel vulnerabilities in ChatGPT Agent systems through multidisciplinary cybersecurity and AI agent security approaches.