AI news · Topic
AI rules, laws and safety
New laws, investigations and industry agreements that change what AI companies are allowed to do, explained in plain words with a link to the original source.
- Four AI giants agree on one way for websites to say no to training (Plagiarism Today, ). Anthropic, OpenAI, Google and Microsoft have announced a shared setting for a website's robots.txt file. It asks their crawlers not to use the site's content to train AI.
- US watchdog looks into OpenAI, Anthropic and other AI firms (Regulators, ). The Federal Trade Commission (FTC), which protects US consumers, is looking into AI companies. It wants to know whether what they say about safety is true, and how they handle AI that acts beyond what users asked.
- California: an AI can no longer fire someone on its own (Regulators, ). California's new “No Robo Bosses Act” says that if a company mostly uses AI to decide to fire or discipline someone, a human must check the decision first. It starts on 1 July 2027.