Morning Edition · Sunday, July 5, 2026
Geopolitics
Anthropic Cuts Off Claude Code in China Over Distillation Claims, Alibaba Bans It Back
The dispute turns a developer tool into the newest front in the US-China split over frontier models.
Tech
Anthropic Ships Claude Sonnet 5, Aimed Squarely at Coding and Agents
The mid-tier model is aimed at frontier performance on software and agent tasks at production scale.
Tech
Meta Says Its Unreleased 'Watermelon' Model Has Caught OpenAI's GPT-5.5
The claim that it matches GPT-5.5 is internal, based on a model still in training, and unverified, and it uses roughly ten times the computing power of Meta's Muse Spark.
Tech
Anthropic Opens Claude Science, a Workbench Built for Running Real Experiments
The app combines researcher tools, records that can be audited, and access to computing power, moving agents into the scientific process.
Tech
Mistral Updates Leanstral, Its Model for Machine-Checked Proofs in Lean 4
Version 1.5 targets formal verification, where a compiler, not a benchmark, decides whether the answer is correct.
Tech
Meta's Brain2Qwerty Decodes Typed Words From Brain Signals Without Surgery
The work trades the accuracy of implants for wider access, decoding language from recordings taken outside the body.
Tech
Anthropic Redeploys Fable 5 and Pitches an Industry Standard for Scoring Jailbreaks
A shared system for rating severity, backed by Amazon, Microsoft and Google, aims to make the risk from jailbreaks comparable across labs.
Tech
Meta's Segment Anything Powers a Promptable 'Digital Closet'
A consumer fashion app shows flexible image segmentation moving from a research suite to a shipped product.
Tech
AI Data Centers May Use Far More Water Than Big Tech Reports, WSJ Finds
Disclosure gaps depend on how a facility is powered, and most providers report only direct water use on site.
Tech
Meta Details How It Builds and Tests Advanced AI at Scale
An engineering account of the evaluation and release systems behind frontier models, not the models themselves.
Tech
Hinton Argues AI Already Learns More Efficiently Than the Human Brain
The claim, from one of deep learning's founders, is about learning efficiency, not total capacity.
Tech
Google's Health AI Portfolio Signals the Push Toward Clinical Deployment
A dedicated research track, not a single product, reflects how labs are structuring medical AI for real care settings.