Bill Gates wants safeguards against an AI catastrophe as his Epstein ties continue to haunt him
His call for stronger oversight comes after a review found repeated warnings about a relationship that damaged trust
B ill Gates wants decisions about artificial intelligence to involve more than the companies building it. His message is straightforward: when technology carries potentially catastrophic risks, safeguards should be mandatory. That argument takes on another dimension after a review found that his foundation had repeatedly raised concerns about Jeffrey Epstein.
In a September 27 interview with NBC’s Meet the Press , the Microsoft co-founder warned that AI could contribute to events causing a billion deaths. He called for lawmakers and law enforcement to help determine how the technology should be monitored. The figure described a possible catastrophe, not an inevitable outcome or a forecast supported by a probability estimate in the remarks cited.
The warnings Gates had already received
A July review by law firm WilmerHale found that Gates knew of reputational concerns surrounding Epstein’s previous sex-related conviction when their discussions began in 2011. Foundation employees raised the risks of the association on multiple occasions, including with Gates.
Investigators identified roughly 30 meetings involving Epstein and ten foundation leaders and employees, Gates among them, rather than 30 meetings attended by Gates alone.
Those discussions largely concerned philanthropy, including a proposed fund intended to attract wealthy donors to global public health. According to the review, the foundation abandoned that proposal after a December 2014 meeting made clear that Epstein had misrepresented potential donors’ knowledge and willingness to participate. The promised funding partnership never materialized.
The investigation found no evidence of participation in or knowledge of Epstein’s sex trafficking operations, or illegal conduct in the meetings examined. Its scope did not include Gates’s personal communications or private dealings outside the foundation.
The board subsequently approved stronger procedures for vetting partners, escalating risks and managing conflicts of interest.
That history gives Gates’s latest appeal a sharper question: what happens when warnings reach powerful people but fail to change their decisions?
His foundation’s experience illustrates why having concerns on record is only a first step. For AI, the practical issue is whether the safeguards he advocates would give someone the authority to intervene before a dangerous decision becomes an irreversible one.
