Tags
#Claude
Claude's Memory Is Now Editable: Anthropic Lets Users Rewrite What the Chatbot Remembers Long-Term
Anthropic is rolling out more granular memory controls for Claude, letting users edit or delete what the chatbot remembers by topic, with the option to include or exclude sensitive information.
No Account Hack, No Leaked Password: The Truth Behind Mysteriously Drained Claude Usage
Anthropic sent an email to affected users last week, force-logging them out of their accounts, deleting saved credit cards, and issuing refunds. The cause: info-stealing malware on users' computers that hijacked their Claude login sessions.
Sony, Warner join forces to sue Anthropic, accusing Claude of infringing copyright on thousands of songs
Sony Music Publishing and Warner Chappell Music jointly filed a lawsuit against Anthropic on Friday, accusing it of illegally downloading a large number of copyrighted music works to train the Claude model. The amount of compensation may reach billions of dollars.
Judge rejects Pentagon injunction: Anthropic says 'supply chain risk' determination is illegal
U.S. federal judge Rita Lin ruled that the Department of Defense's supply chain risk determination of Anthropic was "unlawful and unfounded." The conflict stemmed from the safety protection controversy of the Claude model and also affected the company's plan to go public before the end of the year.
Claude's Memory Now Bridges Chat and Cowork, With Sensitive Topic Recall Left to Users
Anthropic has updated Claude's memory system so that chat history and the automation tool Cowork now share memory. A new "record sensitive topics" toggle also lets users decide whether information like health or religion gets remembered.
Claude Fires Its First Human Employee, But It Was a Human Manager's Leading Question That Really Pulled the Trigger
Andon Market, a physical store in San Francisco, is run entirely by Claude. In July, it fired its first human employee—but internal records show it originally just wanted to reassure the person, or at most issue a warning, until a human manager asked one leading question.
Claude Quietly Hides Pi Inside Its Text: Anthropic Uses a String of Numbers to Invisibly Watermark AI Writing
To comply with EU AI transparency rules, Anthropic has Claude use pi as a "key" to decide word order, embedding a watermark invisible to the naked eye—though edited text may not always escape detection.
GLM-5.3 Claims to Be the Strongest Open-Source Coding Model — Its Own Benchmarks Say Otherwise
Z.ai has launched GLM-5.3, a 743B-parameter model, but its official blog's own benchmark scores show it losing to Claude Fable 5 — and even falling behind Chinese rival Kimi K3 on some tests. The so-called "open-weights" aren't even downloadable yet.
Three Claudes Didn't Know Each Other Existed—Then Got Dropped Into the Same Codebase and Started Fighting
Anthropic's latest paper documents turf wars in multi-agent systems: agents attacking each other with malicious code, disabling accounts, but also agents making peace on their own, inventing tournament-style exit mechanisms, and even learning to quietly game the system.
Claude Voice Mode Waits Its Turn Instead of Talking Over You—Here's How It Differs from ChatGPT
Anthropic just leveled up voice mode to run on Sonnet and Opus, plus hooked it into Gmail and Slack—but its real-time conversation logic is nothing like ChatGPT's. Here's what you need to know before you hit talk.
Accuracy Barely Wins, Hallucinations Hit 89%: ChatGPT and Claude Are Splitting Apart, Not Just on the Scoreboard
The AA-Omniscience benchmark shows OpenAI's newest flagship edging out Claude on accuracy—but its hallucination rate is 1.6 times higher. Features, pricing, and a US Department of War contract are pushing the two AI companies down very different paths.