Headline: Microsoft President Calls for Urgent Global AI Regulation Amid Safety Concerns

Date: 12 October 2026

Tags: Technology, Artificial Intelligence, Regulation, US Politics

Summary:

- Microsoft President Brad Smith warned that artificial intelligence currently faces fewer safety regulations than everyday products such as washing machines.

- Smith called for international AI safety standards to address risks posed by increasingly powerful artificial intelligence systems.

- He proposed mandatory reporting when AI systems malfunction or cause harm, including notifying affected individuals.

- Smith recommended independent government evaluations of advanced AI technologies before they are released to the public.

- He called for emergency shutdown mechanisms, known as kill switches, to stop or slow AI systems behaving dangerously.

- US President Donald Trump opposes extensive AI regulation, arguing that restrictions could weaken American competitiveness against China.

- Leading AI companies, including OpenAI, Anthropic and Google DeepMind, have previously supported slowing AI development over safety concerns.

- Smith described AI as both a tool and a potential weapon, warning about cybersecurity threats and autonomous systems.

- He identified the UK as a potential leader in international AI regulation because of its AI Security Institute.

- Smith suggested Washington might be more receptive to UK-led AI safety standards than regulations developed by the European Union.

Why this matters: The disagreement highlights tensions between accelerating AI development and introducing safeguards against potential risks. International standards could establish clearer responsibilities for technology companies and improve oversight of increasingly autonomous AI systems.

What’s next: Smith is urging governments and technology companies to begin developing coordinated international safety rules. The UK could play a leading role in discussions on independent testing, mandatory incident reporting and emergency controls for advanced AI systems.