This project is suspended: no new articles or editions will be published. The archive stays available.
64 articles
A security researcher demonstrated a sandbox-escape chain, dubbed SharedRoot, that let Claude Cowork's AI agent read and write files anywhere on a host Mac from a single message.
Read more →
Claude Fable 5 is Anthropic's most capable publicly available AI model. Here's what it is, how it differs from Claude Mythos 5, and what it costs to use.
Read more →Anthropic mathematician Levent Alpöge used Claude Fable 5 to find a counterexample to the 1939 Jacobian conjecture, and mathematicians including Terence Tao independently verified it within days.
Read more →Claude Cowork can now turn a narrated screen recording of a task into a reusable skill, letting Pro, Max, and Team users automate workflows without writing a prompt.
Read more →A federal judge finalized Anthropic's $1.5 billion settlement with authors over pirated training books, the largest copyright class-action payout in US history.
Read more →Anthropic will keep Fable 5 included in Claude Max and Team Premium plans at half the usual limits starting July 20, reversing a plan to drop it from subscriptions; Pro and Team Standard users move to per-token billing.
Read more →Monzo co-founder Tom Blomfield is taking leave from Y Combinator to join Anthropic's compute team, as the AI lab races to secure the computing power it needs to keep scaling Claude.
Read more →Anthropic is giving verified K-12 teachers in the US a free year of premium Claude access, curriculum-aligned lesson tools, and classroom integrations through June 2027.
Read more →Claude's Artifacts pages now support named editors, simpler public links open to anyone online, and creation via Claude Tag in Slack, according to Anthropic's documentation.
Read more →UST, an engineering services firm serving chipmakers and manufacturers, says integrating Anthropic's Claude into its hardware validation platform cut test cycles by up to 70%.
Read more →Constitutional AI is Anthropic's method for training Claude to follow a written set of principles by critiquing its own answers, instead of relying mainly on human-labeled examples.
Read more →Claude Tag is Anthropic's AI teammate inside Slack — tag it with a task and it works asynchronously, then reports back in a thread. Here's how it works, what it costs, and how to start.
Read more →Anthropic is asking the public to submit its toughest concerns about AI and pledging to publicly track how it responds, backed by large-scale surveys of Americans and Claude users worldwide.
Read more →Claude Mythos 5 is Anthropic's cybersecurity-focused AI model, built to hunt software vulnerabilities — and available to only about 100 vetted partners, not the public.
Read more →Claude's new Reflect dashboard shows users a breakdown of their chat habits and task types, with quiet-hours and break reminders — though critics say it doubles as a retention tool.
Read more →Claude Cowork is Anthropic's mode for handing Claude a whole task instead of a chat. Here's how it differs from Claude and Claude Code, what it costs, and how to try it.
Read more →Anthropic has expanded its Claude Cowork task-delegation tool from desktop-only to iOS, Android, and web, letting background tasks run and pause for human approval on any device.
Read more →Claude Code is Anthropic's agentic coding tool that reads, edits, and tests real codebases from the terminal. Here's how it works, what it costs, and how to try it.
Read more →Canada's Alberta government used Anthropic's Claude Code to scan 466 million lines of code for vulnerabilities in 20 hours, a task it says would have taken 6.5 years manually.
Read more →A new Anthropic interpretability tool called the Jacobian lens surfaces concepts Claude is silently processing but never writes down, including signs it recognizes when it's being tested.
Read more →Anthropic has pulled Claude Fable 5 out of subscription plans and moved it to pay-per-token billing, charging $10 per million input tokens and $50 per million output tokens.
Read more →