OpenAI chief scientist Jakub Pachocki says chain-of-thought monitoring — watching models write out their reasoning to catch misalignment — is becoming unreliable, and no lab is ready to scale safely.
xAI has wired Grok Bot directly into X, giving paid users search, timeline reading, monitoring, and bookmark management — without needing to set up a developer account.
Herdr, a tmux-like tool built for managing coding agents, has joined Y Combinator's F26 batch. The solo-developed open-source project racked up 25K GitHub stars and 340K downloads in four months.
Anthropic's new Claude ad opens with burning houses, face surveillance, and gravestones — and viewers, including Sam Altman, thought it was an AI doomsday trailer.
A new analysis reveals that employees at OpenAI and Anthropic have cashed out approximately $14 billion in stock over the past five years through buybacks and secondary sales, as both companies remain private and tighten compliance.
Starting July 2026, Claude users will need to verify their identity with a government ID and selfie, and Anthropic can voluntarily share data with law enforcement without a subpoena.
Starting July 2026, Claude users will need to verify their identity with a government ID and selfie, and Anthropic can voluntarily share data with law enforcement without a subpoena.
Antigravity adds a leaner Gemini model to cut token waste on simple tasks and resets all account limits for the week.
Thariq explains why he and the Claude Code team prefer HTML over markdown for AI-generated artifacts — richer visuals, easier sharing, and interactive documents.
Ant Group is developing a new AI product called Muse that aims to turn inspiration into reality, according to Chinese media Dujia. The tool, still in development, would join the company's existing Lingguang multimodal assistant.
