Bill Gates warned in a Sept. 27 Meet the Press interview that artificial intelligence could be weaponized by actors with harmful intent to cause mass casualties on an unprecedented scale. Gates stated that AI is powerful enough to drive events causing a billion deaths, describing the risk as arising from the combination of increasingly capable tools and people seeking to misuse them.
Gates identified biological and cyberattacks as immediate concerns, contending that AI has crossed dangerous thresholds in both areas this year. He noted that much AI development has taken place in the private sector, leaving government with limited technical expertise to manage the risks.
Calls for Mandatory Safeguards and Monitoring
Gates argued that industry self-regulation is insufficient and called for legislation mandating AI safeguards. He advocated for law enforcement and lawmakers to help determine what protections should include, stating that a kill switch alone would not prevent attacks. Instead, he said monitoring systems and record-keeping would be necessary to show what AI systems do and preserve evidence for investigators.
Gates acknowledged that compliance requirements would add work for developers but maintained the burden would not dramatically slow progress.
International Cooperation Remains a Challenge
Cooperation across borders would be necessary to manage AI risks, according to Gates. He compared the challenge to Cold War-era nuclear arms negotiations, then described reaching an AI agreement as even more difficult. Unlike nuclear weapons, which require access to fissile materials, AI systems are more widely accessible, potentially encouraging rapid development without adequate safeguards if treated primarily as national competition.
On Sept. 22, United Nations Secretary-General Antonio Guterres similarly called for a multilateral AI risk management framework with independent oversight.
Broader Security Concerns
More than 40 bitcoin and digital asset organizations have requested that leading AI developers provide vetted security researchers with controlled early access to advanced systems. The request reflects concern that attackers could use AI capabilities to identify software vulnerabilities while defenders lack comparable tools.
Fraud using AI-generated deception already poses risks to consumers. The FBI has warned of scams in which people posing as federal agents use AI-generated videos and fake websites to target previous fraud victims. Cryptocurrency holders face similar threats through phishing messages and fake exchanges designed to compromise passwords and recovery information.
