Microsoft President Brad Smith calls for global AI regulation — Cyprus Mail

Microsoft President Brad Smith called for the urgent introduction of global regulation of artificial intelligence, including independent testing of new developments. He said that AI systems currently face fewer safety requirements than milk or household appliances, Cyprus Mail reports.

Safety proposals

In an interview with the BBC published on Monday, Smith said that a set of international safety rules is needed. According to him, if a system is compromised by an uncontrolled AI agent, the affected party should be informed, and there should also be an emergency shutdown mechanism.

The head of Microsoft also supported a legal obligation for companies to prepare incident reports. He stressed that safety rules and inspections already apply to other products, so this experience should be used to regulate artificial intelligence.

More current news is available on the UA.News Telegram channel Telegram.

The United Kingdom’s approach

Smith called the United Kingdom a country with good prerequisites for leadership in AI regulation, particularly thanks to the AI Safety Institute established under former Prime Minister Rishi Sunak. Under voluntary agreements with developers of large models, the institute receives access to them before deployment to assess capabilities and risks.

Technology companies, including Anthropic and OpenAI, have already begun reporting uncontrolled or unwanted behavior by AI agents. UK AI Minister Kanishka Narayan told Reuters that the country would consider regulating advanced models if the current approach proves insufficient.