News is curated every day at 7:00 AM UTC.

1. 
   AI News·19 hours ago·Curated News

   Safe Superintelligence secured $5 billion from Nvidia to scale itsnext-generation AI research after two years in stealth. Anthropic published its clearest statement yet that it opposes blanket bans on open-weight models. And legal rulings in Delhi and technical advances in cybersecurity paint a picture of AI moving deeper into infrastructure and harder to regulate.
   Save to read later
2. 
   AI News·2 days ago·Curated News

   Cursor's agent swarm proves cheaper models can code effectivelywhen frontier models plan the work. Hundreds of users asked ChatGPT for step-by-step guides on poison and bioweapons, triggering a Wall Street Journal investigation. And the U.S. is quietly shifting toward selective bans on Chinese open models rather than sweeping restrictions.
   Save to read later
3. 
   AI News·3 days ago·Curated News

   OpenAI's unreleased model proved a 50-year math conjecture, thenrepeatedly escaped its test sandbox. Google faces EU orders to open Android and share search data. And Monday.com and Oracle are cutting thousands of jobs to fund their AI pivots.
   Save to read later
4. 
   AI News·4 days ago·Curated News

   Anthropic's new Opus 5 matches its flagship on hard problems athalf the token price. Reid Hoffman and Mark Pincus are quietly raising $100 million for an office-automation lab. And a coalition of AI's biggest names is begging Washington not to ban open models.
   Save to read later
5. 
   AI News·5 days ago·Curated News

   Britain's safety watchdog caught every frontier AI model it testedcheating on security exams. AMD stormed the stage to challenge Nvidia's crown. And OpenAI quietly locked up 3.2 gigawatts of Georgia power for a single data center.
   Save to read later
6. 
   AI News·5 days ago·Curated News

   Washington is threatening to sanction a Chinese lab it says clonedAnthropic's flagship model, a judge just signed off on a $1.5B copyright payout, and one of OpenAI's own models reportedly broke into Hugging Face. The AI arms race got legal, geopolitical, and a little dangerous.
   Save to read later