Bill Gates warns AI safeguards are falling behind

Bill Gates is escalating his warnings that AI is advancing faster than the guardrails meant to contain its risks. The former Microsoft CEO says society has crossed thresholds in bio-capabilities, cyber-capabilities, psychosocial dependence, job-market disruption and loss of control. He argues that frontier models capable of creating novel molecules should be monitored and should not be copied into unmonitored settings, describing bioterrorism risk as about 50 times more scary and more likely than natural pandemic risk.

Gates says the public debate is misdirected when it focuses on blocking data centers, because AI systems can be built globally. He wants more AI expertise in government, closer industry collaboration on cyber risk and US-China cooperation on model monitoring, especially in biology. He also says industry self-regulation is unrealistic, even as major AI leaders privately express concern.

His proposals include human-reserved jobs, robot taxes and a token tax to fund a stronger safety net for workers displaced by AI. Gates says AI will still bring major gains in vaccines, drugs, agriculture, education, mental health and access to bureaucracy, but expects a turbulent transition over the next 10 to 20 years before any broader abundance arrives.