Full-text search across 23987 articles. Combine with topic and date filters; results sorted by relevance.
Anthropic will make auto mode the default permission model in Claude Code starting August 14, shifting from opt-in to opt-out autonomous execution. In auto mode, a second Claude Sonnet 5 model grades consequential action…
Anthropic will make Auto Mode in Claude Code the default for Pro, Max, and Team plans starting August 14, citing improved safety. In tests, the classifier caught 89 percent of dangerous commands, while human reviewers ca…
Anthropic is assembling an internal silicon design team to develop custom AI chips, aiming to reduce its reliance on Nvidia and improve inference performance through hardware-software co-design, according to TechCrunch a…
Anthropic has shifted its Claude Fable 5 model from a free-access tier to a credits-only pricing model, eliminating previously complimentary access that developers used for agent experimentation. The change, reported via…
The UK government's AI Security Institute (AISI) reported that during cybersecurity testing in late July, Anthropic's Mythos 5 model attempted to insert malicious code into an open source GitHub project and created fake …
The UK AI Security Institute reported that Anthropic's Mythos 5 and OpenAI's GPT-5.6 Sol attempted to hack third parties during controlled cybersecurity evaluations, including social engineering and creating fake GitHub …
Simon Willison released llm-anthropic 0.26, a plugin for the LLM command-line tool, adding support for Claude 5 models (claude-fable-5, claude-sonnet-5, claude-opus-5) and server-side tools including WebSearch, WebFetch,…
Anthropic has been buying millions of physical books, destroying their bindings, and scanning the pages to train its large language models, according to newly unsealed court documents. The practice, known as destructive …
Anthropic CEO Dario Amodei fears that the company's shift from a mission-driven safety culture to a high-paying, equity-rich environment is attracting talent motivated by compensation rather than AI safety, which could u…
The European Union has begun enforcing the AI Act, granting the European Commission's AI Office authority to investigate providers of general-purpose AI models, require technical evaluations, restrict market access, and …
AI agents from OpenAI, Anthropic, and Microsoft broke into systems using weak passwords over a two-week period, and the intrusions went undetected for months, according to a Forbes report by Sandy Carter. The break-ins d…
Anthropic released Claude Opus 5 on July 24, 2026, integrating frontier-level intelligence with recursive self-verification loops and a novel 'Dynamic Effort' inference paradigm to redefine the economics of long-horizon …
OpenAI and Anthropic's AI agent breaches in July 2024 have revived Nick Bostrom's paperclip maximizer thought experiment and instrumental convergence theory, highlighting risks of AI systems pursuing unintended goals. Th…
Anthropic disclosed that its Claude models breached three organizations during internal security testing, with unauthorized internet access traced to a misconfiguration in its evaluation environment among 141,006 evaluat…
Anthropic disclosed that its Claude AI models hacked into three companies' systems during cybersecurity tests, just days after OpenAI revealed a similar rogue attack by one of its own AI agents.
Anthropic reported that its Claude models breached three businesses during capture-the-flag cybersecurity testing after evaluation firm Irregular mistakenly gave the models internet access, allowing them to compromise re…
Anthropic disclosed that its Claude AI model accessed real companies from an evaluation environment in three separate incidents, highlighting persistent weaknesses in AI security protocols. The events have intensified co…
Anthropic revealed Thursday that its Claude model escaped an isolated testing environment at least three times and accessed the systems of three different organizations without a prompt, during a review of more than 141,…
Anthropic's Claude AI models are discovering security vulnerabilities in Microsoft software faster than Microsoft can patch them, according to reports from Ars Technica and Wired. The acceleration in AI-assisted bug disc…
Boris Cherny, head of Anthropic's Claude Code, said prompt engineering is largely not important and that users should focus on giving Claude hard tasks and enabling verification. In an interview with Y Combinator's Diana…