Gates warns of imminent AI-driven cyber and bio threats

Bill Gates has identified cybersecurity and biosecurity as the two most significant near-term risks associated with advanced artificial intelligence. In the cybersecurity domain, Gates warned that AI models are capable of identifying flaws in complex software systems used by hospitals, electric grids, and banks. He noted that attackers could exploit these vulnerabilities to shut down or scramble essential services. Gates emphasized that using open-source AI tools with monitoring safeguards removed allows red teams to successfully execute these attacks. He estimated that replacing vulnerable software would take approximately five years, even if the effort were prioritized immediately. Consequently, he predicted a dramatic increase in cyberattacks over the next few years, driven by the established business model of ransomware.

In the biosecurity sector, Gates stated that AI could be used to create pathogens significantly more dangerous than the virus that caused the COVID-19 pandemic. Unlike cyberattacks, however, there is no clear business model for biological attacks, which makes predicting their timing and nature more difficult. To mitigate this risk, Gates called for massive detection capabilities in air, sewage, and blood to catch infections early. He argued that such surveillance is both feasible and cost-effective. Gates maintained that the offensive capabilities of AI in both cyber and bio domains require urgent defensive investment and global surveillance efforts to protect critical systems and public health.