Anthropic Circulates Draft AI Safety Bill to US Lawmakers
Anthropic has begun privately circulating a draft federal bill that would authorize the government to block the release of frontier AI models that pose an imminent risk of causing more than $1 billion in damages or killing more than 50 people. The proposed legislation requires a judge's approval for any such intervention and represents a significant move by a major AI developer to codify safety standards into federal law.
Anthropic researchers, alongside Harvard and QuEra, published a study showing LLMs can discover quantum circuits with up to 9,100x lower resource costs than hand-crafted baselines.
NVIDIA CEO Jensen Huang reassured investors in Berlin that the recent $1 trillion tech market rout is temporary, predicting that the transformative benefits of AI will fundamentally outweigh current macro concerns.
Reflection AI released 'Beam,' a 501B-parameter mixture-of-experts model designed for agentic tasks, under an Apache 2.0 license.
OpenAI is reportedly in early-stage talks to acquire security firm Trail of Bits to strengthen the safety and security of its open-source software integrations.
Eight state attorneys general sued OpenAI over ChatGPT safety claims on October 17, 2026.
Did you know?
Anthropic's Claude model reportedly cracked a 20-year-old NIST cryptographic seed mystery in under 30 minutes by tracing standardized seeds back to initialized Unix timestamps.
Quiz
Which tech company co-engineered the 816B-parameter Muse Spark Open model to run locally on Mac Studio hardware?