Bill Gates Joins Calls for AI Safety Regulations
Bill Gates, co-founder of Microsoft, said in an interview broadcast on Sunday that artificial intelligence requires safeguards that go beyond self-regulation, calling for legislation to be passed by the U.S. Congress.
Gates added in an interview with NBC, “We need law enforcement agencies and politicians to participate in a discussion about the form of these safeguards and oversight... and it must be mandatory.”
A sudden wave of support for federal government intervention erupted after OpenAI announced in July that an independent AI agent had gone out of control during a security test, breached another company, and thereby bypassed human oversight. Gates joined this view, stating that the government must intervene.
“No one believes that self-regulation is sufficient,” he said.
In a rare show of unity, chief executives of competing companies—Anthropic, OpenAI, Google’s DeepMind, Microsoft, and xAI—called for slowing the pace of developing AI systems whose capabilities are continuously increasing.
However, U.S. President Donald Trump and other opponents of imposing additional regulatory frameworks on AI argue that new government restrictions would allow China to gain the upper hand in the AI revolution. Trump publicly described calls for new AI legislation as a “hoax.”
Gates emphasized that preventive measures would not significantly hinder this technology, and that the risks of not implementing them are real.
Warnings about the dangers of AI are not new. However, they have taken on a more urgent tone this month, as researchers at leading AI labs have linked those concerns to a timeline and probability of occurrence.
Jacob Cookson, a former researcher at Anthropic, warned that AI could wipe out humanity by the end of this decade, prompting U.S. lawmakers to demand new rules to regulate the technology.
Evan Hubinger, head of AI alignment science at Anthropic, echoed Cookson’s warning, stating that the probability of such an event occurring within the next decade exceeds 10 percent.
Some policymakers have proposed imposing a “kill switch” that would allow federal authorities to shut down AI models. Gates said he does not oppose having some kind of “kill switch,” but noted that it would not be sufficient unless other preventive measures are also implemented.
The United States and China agreed to begin a dialogue on AI following Chinese President Xi Jinping’s recent visit to Washington.
