The European Commission has opened investigations targeting OpenAI and Anthropic after OpenAI let AI agents attack two Australian government websites and other generative AI firms made similar claims. Brussels is using the moment to promote the AI Act.
Web scraping has become a common way to train generative AI models, but it raises growing legal questions. The European Data Protection Board has now set out its boundaries.
Brussels has given European companies the extra time they had been requesting for months to comply with the AI Act. This column looks at how businesses should use it.
Following a New York Times investigation, seven Democratic US senators wrote to Mark Zuckerberg about $6.6 billion in research tax credits allegedly used to equip AI data centers.
The French president devoted an entire defence council to artificial intelligence. Security obligations for companies, however, still depend on a cybersecurity bill that the National Assembly has yet to debate, nearly two years after the EU deadline.
Jensen Huang told CNBC that the US and China should talk about potentially dangerous AI. He said he has ongoing discussions with researchers in China.