Anthropic removed an undisclosed telemetry feature from its Claude Code CLI after users and security researchers flagged background data transmissions lacking clear disclosure.
Read more →
Anthropic's Claude Cowork launches in beta on web and mobile for Max subscribers, with a Dispatch feature for cross-device task continuity and new usage data.
Read more →
Meta's Muse Image generator is free in the Meta AI app, Instagram, and WhatsApp, but a feature that pulls in tagged users' photos without notice has critics calling it a privacy landmine.
Read more →
xAI is reportedly preparing a coding-focused AI model with Cursor integration, though key details remain unverified as SpaceX, xAI, and Anysphere have not commented.
Read more →
Thinking Machines AI found that, with expert human annotations, language models can be trained to filter financial documents with investor-level taste and judgment.
Read more →
OpenAI launches GPT-5.6 on July 9, 2026, with three tiers (Sol, Terra, Luna), new reasoning modes, and pricing from $1/$6 to $5/$30 per 1M tokens.
Read more →
Anthropic gives Claude Pro, Max, Team, and select Enterprise subscribers until July 12, 2026 to use Fable 5 free, up to 50% of weekly usage, before paid credits take over.
Read more →
Varonis Threat Labs found a code injection flaw, dubbed Rogue Agent, in Google Dialogflow CX that let attackers exfiltrate chats and run phishing at scale.
Read more →
Claude Tag lets Slack teams mention @Claude to delegate tasks like summaries, drafts, and follow-ups. Here is the setup, permissions, and cost-control checklist.
Read more →
Cursor's AI code editor is now on iPhone and iPad, with touch-optimized editing, background coding agents, and synced desktop sessions.
Read more →
After the Mythos moment, AI policy fights are reshaping regulation: US-China gaps, DeepMind worker leverage, the AI Incident Reporting Act, and the courts stepping in.
Read more →
Anthropic restored Claude Fable 5 worldwide a day after the U.S. lifted export controls, following a new safety filter reviewed by Commerce.
Read more →
OpenAI is preparing a sweeping ChatGPT redesign into a superapp combining chat, search, and tools, planned alongside a potential IPO, per the FT.
Read more →
Google has published a Gemini 2.5 Computer Use model in the Gemini API that drives browsers and UIs from screenshots and natural-language instructions.
Read more →
Meta's Brain2Qwerty v2 reads intended text from brain activity using a non-invasive MEG helmet and deep learning, reaching 61% word accuracy.
Read more →