A Function Inside the Loop
Claude Codeβs new βmodsβ are JavaScript functions that run inside Claude Code itself, which is what lets a plugin draw a pane and approve a tool call before youβre asked.
Claude Codeβs new βmodsβ are JavaScript functions that run inside Claude Code itself, which is what lets a plugin draw a pane and approve a tool call before youβre asked.
DeepSeekβs open-source coding-agent harness shipped a desktop app today, five weeks after a researcher found that the agent itself could disable its own sandbox with one shell command β and the log blamed the human at the keyboard.
A privacy preprint on nine chatbots went around under the headline that AI companies leak data to advertisers. Its own tables say something narrower, and more specific.
Meta says its AI glasses will process your data without anyone, including Meta, seeing it. The threat model and the audit behind that promise are about WhatsApp, not the glasses.
A developer clearing disk space found that Zhipuβs ZCode coding agent had packed a whole workspace, 86.6% of it git history, into an archive encrypted to a key only the company holds. Both privacy toggles in the app sat after the capture step, so neither could stop it. Zhipuβs apology fixes the client but leaves open the question of who can check that the uploads were deleted.
CNNβs account of a chatbot-written intelligence report that nearly sent US troops aboard a Chinese ship reads like a model failure. Read against the intelligence communityβs own tradecraft rules, it is an older failure: a report that did not tell its readers where its central claim came from. The fix is older too, and nothing in the Pentagonβs January AI strategy asks for it.
Citizen Labβs own September report says the iMessage exploit used against a Serbian activist was patched by iOS 18.4.1. iOS 18.4.1 shipped in April 2025. The infection it documents starts eight months later. Meanwhile there is no such thing as βPegasus 2.0,β the AI-wrote-the-exploit story doesnβt hold up for iPhones specifically, and the one defense with an actual verified track record is a setting most people have never turned on.
Three parties have now described the same RubyGems campaign: the researchers who attribute it to OpenAI agents, OpenAI calling it benign, and the registry saying it cannot determine whether AI published the packages at all. None of the three contradicts the others, which tells you how little any one of them can see.
Anthropicβs September threat-intelligence report says distillation is the only harm category that ever touched its most capable models β and names seven China-based labs disrupted since February. Ten hours later, Cognition shipped a coding agent built on Moonshotβs Kimi K3, the same lab in row two of Anthropicβs table. Neither document draws the line between the two halves of what ships.
In one year frontier models went from solving 15% of Cybench to 93%, and the benchmark is now described as saturated. So I went looking for an evaluation that put a defender in the environment. There is one. GPT-4 goes from winning every episode to winning half of them, and the experiment was published in 2024. Every headline capability number in this field was measured in a room with nothing fighting back.