“Cyberattacks” are the most immediate AI danger, Bill Gates warns, urging safeguards as he rejects relying on companies to police the technology

Microsoft co-founder Bill Gates warned Sunday that malicious users exploiting artificial intelligence without safeguards pose a more immediate danger than AI escaping human control, and predicted a sharp rise in AI-powered cyberattacks over the coming year. He made the remarks in an interview with ABC News, AFP reported.

Gates said the risk of AI getting out of control with major consequences remains years away. For now, he said, users can remove safety restrictions, and combining bad actors, unguarded systems and the latest models could bring “a lot of negative surprises,” particularly in cyberattacks.

He called for stricter regulation to prevent malicious people from using AI for bioterrorism and cyberattacks. Gates also rejected the position that AI companies’ own oversight is enough, arguing that relying on the companies as society’s only check on “the most dangerous technology humans ever invented” is inadequate.

On September 29, President Donald Trump and representatives of major AI companies signed a voluntary accord focused on companies’ internal controls and external reviews. Trump said the industry should rely on self-regulation rather than broad government rules.