Bill Gates is preparing to significantly expand his efforts on AI safety, personally engaging business leaders and potentially deploying substantial resources, according to sources told to Axios. The shift marks a return to the competitive intensity that characterized his early leadership of Microsoft, now directed toward what he views as a critical global threat.
Gates has moved from a stance of measured optimism to one of urgent concern, driven by escalating model capabilities and alarming security incidents. Sources indicate he believes current regulation is insufficient. To build his case for a more robust regulatory framework, he consulted extensively with key figures in the industry, including Anthropic CEO Dario Amodei, along with other researchers and safety advocates.
Associates describe this advocacy as a "comeback tour," aimed at rebuilding a legacy tarnished by his past association with Jeffrey Epstein. Gates acknowledges that these ties may lead some to dismiss his warnings. In an August interview with Axios, he admitted that his previous engagement with Epstein was a "huge mistake" motivated by a desire to raise money, though he stated the stakes of AI risks are too high for him to remain on the sidelines.
His outlook has darkened sharply over the past three years. In 2022, Gates argued that AI risks were real but manageable, comparing the technology's disruption to the introduction of the PC or the internet. By August of last year, he had reversed this view. In a lengthy essay and subsequent interviews, he warned that the current trajectory is volatile and that society is dangerously unprepared for the upcoming economic and societal upheaval.
Gates has not yet defined the exact contours of his campaign, but his research has led him to advocate for specific policy changes. In his August essay, he called for new regulatory institutions to coordinate AI policy, mechanisms to reserve certain jobs for humans, and taxes on AI usage and robots. He believes his network of contacts, technical credibility, and wealth can be leveraged effectively to influence these outcomes.
This approach contrasts with the softer philanthropist persona he has cultivated over the last two decades. In recent interviews, he has displayed a sharper edge. He recently chided journalist Ezra Klein for suggesting that normal corporate reputational management might suffice for regulation. Gates sarcastically suggested that lawsuits should not be the primary regulatory tool, remarking on the absurdity of using legal action in a hypothetical scenario involving mass casualties.
His push for robust regulation places him in direct conflict with the Trump administration, which has favored a lighter regulatory touch. The administration has organized a group of top officials to police the industry through a voluntary safety pact rather than binding legislation. As Gates' plans take shape, it remains unclear whether AI safety will become the defining battle of his late career.