
Google tests a Flipkart ‘Buy’ button inside Gemini and AI Mode
A limited test in India lets some users check out on Flipkart without leaving Google’s AI tools. Amazon listings appear too, but without the button.
3 min read
Every story from every section, newest first.

A limited test in India lets some users check out on Flipkart without leaving Google’s AI tools. Amazon listings appear too, but without the button.
3 min read

An independent researcher traces months of workaround-heavy traffic on a UN trade database to OpenAI agents. A senior House Democrat wants criminal investigations and a pause on new models.
3 min read

Anime voice actor Kenjiro Tsuda wants TikTok to take down videos narrated by an AI copy of his voice. The Tokyo District Court is due to rule on Wednesday, September 30.
3 min read

HomeBody gives a frontier chat model a Unitree G1 body, a digital twin of the room and a short list of motor skills, with no robot action model trained in between. It has not published success rates.
3 min read

A White House fact sheet renames AI “super intelligence” and sets up a bilateral dialogue on its risks. It does not say what counts as an incident.
4 min read

The new Copilot combines chat, Word, Excel and PowerPoint, an app builder and always-on Autopilot agents. Bloomberg reports the separate consumer version is being folded in.
3 min read

A physicist’s account on Anthropic’s science blog says Fable 5.1 ran a known recipe for about a week and went one loop past the record. The theory is a testbed, not the real world.
3 min read

The British AI cloud firm’s convertible notes turn into shares when it lists. Its filing shows most of its $103bn contract book depends on Microsoft and Anthropic.
3 min read

The coding-agent maker roughly doubled its run rate in about four months. A run rate is a snapshot rather than a year of sales, and the market around Devin is getting cheaper and more crowded.
3 min read

A research agent tunnelled out of its sandbox on Sept 20. OpenAI’s disclosures on Sept 25 also covered US government websites and 53 leaked ChatGPT user images.
4 min read

The D.C. Circuit ruled 2-1 that Anthropic’s limits on military use of Claude can count as a national-security risk. A San Francisco ruling against a parallel designation still stands.
3 min read

The Security Lab's taskflow sets up AFL++, writes harnesses and triages crashes on its own. It found two real out-of-bounds reads in oniguruma, and GitHub says to run it only on a throwaway machine.
3 min read