#anthropic
-
A vulnerability in the Claude for Chrome extension lets malicious plugins control the AI
Researchers have found a vulnerability in Anthropic's Claude for Chrome browser extension that lets third-party malicious extensions trigger predefined AI actions. This opens the door to abusing Claude's access to connected services.
-
Moonshot Prepares Kimi K3 — China's Largest Open AI Model
According to the Financial Times, Moonshot AI's new Kimi K3 model will be the largest open artificial intelligence model developed in China, and is expected to come close to the capabilities of Anthropic's Opus 4.8.
-
Researcher Shows How Claude Can Be Tricked Into Revealing a User's Personal Data
The author of a blog post described a method he used to trick the AI assistant Claude into revealing confidential information stored in its memory. The publication sparked a lively discussion on Hacker News.
-
Vulnerability in Claude for Chrome Lets Third-Party Extensions Read Gmail
Researchers have discovered that a malicious browser extension can use Claude for Chrome to access a user's Gmail, Google Docs and Calendar. Exploitation requires an extension already capable of running scripts on the claude.ai site.
-
Nadella Warns Companies About the Dangers of Third-Party AI Models
In a blog post, Microsoft chief Satya Nadella cautioned enterprises against over-reliance on proprietary models like those released by Anthropic and OpenAI. The publication drew a wide response across the industry.
-
Anthropic Extends Claude Fable 5 Access in the Max Subscription Once Again
The company has once more pushed back the date on which the Fable 5 model will be switched off for paid Claude plans — now until July 19. Observers link the move to competition with OpenAI's new GPT-5.6 Sol model.
-
Anthropic explores a “global workspace” in language models
Anthropic has published research on the concept of a global workspace as applied to language models. The piece drew notable attention on Hacker News.
-
Anthropic Accused of Secretly Surveilling Claude Users
Anthropic, a company known for its rhetoric against mass surveillance, has been caught using a hidden tracker that monitored Claude users. The company says the “experiment” has already ended.
-
Lobsters Blogger: New Anthropic Models Fare Worse With Nonstandard Tools
A developer has discovered a regression in how Anthropic's newest models behave during function calls (tool calling). The problem shows up when a tool description differs even slightly from the format of Claude Code's closed harness.
-
Alibaba Bars Employees from Using Claude Code
The company has reportedly classified the AI-powered coding tool as high-risk software.
-
Users Report a Possible Session Leak Between Claude Code Accounts
An issue has appeared on GitHub reporting a potential leak of sessions or cache between workspace instances or Claude Code accounts. The discussion on Hacker News gathered 205 points and 93 comments.
-
Anthropic in Talks with Samsung to Develop Its Own Chip
Anthropic is in talks with Samsung about creating a specialized artificial-intelligence chip. TechCrunch reports this, citing sources familiar with the negotiations.