Judge says publishers suing Anthropic cite no infringing Claude output
A judge says music publishers alleging Claude copied lyrics from 20,517 songs have not shown one such output. A status report on the claim is due 13 November.
3 min read
AI Observer: News, guides and analysis on AI agents, workflow automation and chatbots, for teams that want to save time on repeated work.
Nous says its Arete-27B fine-tune cut Qwen propaganda answers from 22% to 0.9%
Nous Research's open Arete-27B retunes Qwen3.8-27B. On Nous's own 803-prompt test, propaganda answers fell from 22% to 0.9%. No outside replication yet.
3 min read
Claude now connects teammates to company apps without a sign-in screen
Anthropic's auto-connect beta signs Team and Enterprise members into 14 apps on first use, with no identity provider needed. Offboarding is the catch.
3 min read
EU court adviser sets a test for when chatbot answers infringe publishers
Advocate General Szpunar says chatbot answers can infringe press publishers' rights after unlawful access or reuse that replaces the article. Judges rule later.
3 min read
NVIDIA’s GR00T 2 learns robot skills from human video, code due December
NVIDIA says GR00T 2 hits 95% on 50 robot tasks after five demos each. Its own whitepaper says that number is high partly by construction.
3 min read
ChatGPT Remix lets Business and Enterprise teams fork a teammate’s site
OpenAI's Remix copies a teammate's ChatGPT Site so you can change it in chat. What it does, who gets it, and what teams should check before copying data.
3 min read
ROSS asks full Third Circuit to rehear first appeals ruling on AI training
ROSS Intelligence wants the whole court to revisit its Westlaw headnotes loss, the first US appeals ruling on fair use and AI training.
3 min read
Latest stories
- Judge says publishers suing Anthropic cite no infringing Claude output
- Nous says its Arete-27B fine-tune cut Qwen propaganda answers from 22% to 0.9%
- Claude now connects teammates to company apps without a sign-in screen
- EU court adviser sets a test for when chatbot answers infringe publishers
- NVIDIA’s GR00T 2 learns robot skills from human video, code due December
- ChatGPT Remix lets Business and Enterprise teams fork a teammate’s site
- ROSS asks full Third Circuit to rehear first appeals ruling on AI training
Eight desks, one place to look
Every story on AI Observer lives on a desk. Pick one to see what is new there.
- Models and researchNous says its Arete-27B fine-tune cut Qwen propaganda answers from 22% to 0.9%3 this week
- Agents and automationClaude now connects teammates to company apps without a sign-in screen5 this week
- Chips and hardwareAMD ships its first Helios AI racks but keeps its GPU revenue to itself1 this week
- Business and startupsTraversal raises $120M Series B at a $1B valuation for AI SRE1 this week
- Policy and regulationROSS asks full Third Circuit to rehear first appeals ruling on AI training4 this week
- Health and scienceHealth AI assistants explained: what they do and where your data goes1 this week
- RoboticsNVIDIA’s GR00T 2 learns robot skills from human video, code due December3 this week
- Design and productChatGPT Remix lets Business and Enterprise teams fork a teammate’s site2 this week
Latest stories
All news
The AI Observer Awards will recognise AI work that was checked, shipped and still holds up a year later.
A planned programme. Nominations, jury and criteria will be published here before anyone is asked to enter.
Planned categories
- AI agency of the year
- Best AI tutorial
- Best interview
- Tool that held up