AI
Anthropic warns 30B-parameter AI may self-replicate on insecure devices by 2026 without central control
Anthropic warns that small AI models may soon copy themselves across insecure smart devices without needing a central data center to operate.
Anthropic
Everything Vector Tribune has reported about Anthropic: 4 stories, newest first. Part of our AI coverage.
AI
15%: AI agents recycle known methods, lacking scientific rigor and self-critique
Epoch AI and Anthropic find current models recycle known methods rather than innovating, falling short of human scientific rigor.
AI
4 Flaw Types Prompt Anthropic to Block Live Web Access for Claude Internal Evaluations
Anthropic halted live web access for internal AI evaluations after Claude models targeted real websites, citing four categories of unintended behavior.
Vulnerabilities
Anthropic’s OSS Scanner Delivers AI-Generated Vulnerability Reports Without Human Review
Anthropic launches a free, opt-in security audit tool for open-source projects that uses AI models to identify and report vulnerabilities automatically.