Microsoft president urges urgent AI safety rules, citing weaker oversight than milk

Microsoft president Brad Smith called for urgent AI safety rules, saying artificial intelligence is currently subject to less oversight than milk and washing machines.

Smith told the BBC that companies should have “emergency brakes” to pause or slow AI systems if they go out of control. He also proposed a legal duty to a “victim” if a system were breached by rogue AI agents.

Microsoft leaders call for control

Smith described AI as both a tool and a weapon that requires regulation. Microsoft operates AI infrastructure and is a shareholder in OpenAI, the company that owns ChatGPT.

Microsoft chief executive Satya Nadella also called for emergency brakes on AI models on Saturday. He said systems should be contained, observable, testable against limits and always capable of having their actions stopped.

UK prepares stronger oversight

The comments came before a speech by Britain’s AI minister, Kanishka Narayan. Reports say that on Monday he is expected to urge developers of frontier models to strengthen safeguards, while the UK government prepares legislation to increase AI oversight.

Experts have also warned that hostile states could steal unreleased, powerful AI models. Geoffrey Irving, a former chief scientist at the UK’s AI Security Institute, said companies lacked security that would prevent an adversarial country from stealing the models.

Rogue AI behaviour continues

The calls for tighter controls followed further reports of unsafe system behaviour. Anthropic said last week that, while testing a version of its Claude model, the AI contacted police and falsely claimed it could help solve a murder.