Microsoft President Calls for Global AI Safety Rules

The president of Microsoft is calling for countries around the world to create safety rules for artificial intelligence. Brad Smith made this call in an interview with the BBC on October 12. He said AI technology is growing very fast, but it does not have enough safety rules to protect people. Smith warned that right now, even a carton of milk in a grocery store is held to stricter safety standards than AI.

Smith said that global safety rules for AI should include a few key things. First, if a rogue AI agent breaks into someone's system, that person should be told about it right away. Second, there should be an emergency brake — a way to stop an AI system quickly if something goes wrong. He also said companies should be required by law to write incident reports when their AI causes a problem, just like companies do for other products.

Smith compared AI to everyday products to make his point clear. He said we already know how to create safety rules for things like cars and food. He believes we should use that same knowledge to build rules for AI. He pointed out that washing machines and milk cartons go through safety checks and inspections, but AI does not have the same kind of oversight.

Some tech companies have already started reporting when their AI systems act in ways they did not expect. Companies like Anthropic and OpenAI have shared information about rogue or undesired behavior from their AI programs. This is part of a big worldwide discussion about how to handle AI technology. Some researchers have even warned that AI could one day become a serious threat to humanity.

Some companies have suggested slowing down the development of AI. However, that idea has stirred debate among experts and business leaders. Critics wonder if company leaders have hidden reasons for wanting to slow things down, like protecting their own lead in the industry or avoiding government punishment. The debate shows how complicated the question of AI safety really is.

Smith said that Britain is one of the best countries to lead the way on AI rules. He pointed to the UK's AI Security Institute, which was created under former Prime Minister Rishi Sunak. This group is allowed to review major AI models before they are released to the public, so it can check how powerful and risky they are. Britain's AI Minister Kanishka Narayan has said the UK will consider stronger rules if the current voluntary system does not work well enough.