Experts Clash Over Claims That AI Could Pose Existential Risk to Humanity

A series of stark claims about the dangers of artificial intelligence has surfaced in recent days, prompting debate among researchers, industry figures and commentators over how seriously the public should take warnings of AI-driven catastrophe, according to a report by The Guardian.

Among the claims examined by the newspaper is an assertion that there is a 10% chance that AI could lead to human extinction or a similarly catastrophic outcome. The Guardian's review also cites a claim that artificial intelligence systems pose a greater threat than nuclear weapons, as well as a warning that a so-called "botnet" of AI systems could threaten the stability of the entire internet.

Not all voices in the debate agree that these warnings are well-founded. The Guardian's report notes that some critics have dismissed the more alarming claims as a "psyop" — suggesting that dire predictions about AI risk are being promoted, deliberately or otherwise, to serve the interests of technology companies rather than to reflect a genuine assessment of danger.

The report frames the dispute as part of a broader disagreement over whether the pace of AI development should be slowed. Advocates of caution argue that the rapid rollout of increasingly powerful AI systems outstrips society's ability to understand or manage the risks involved, while skeptics contend that apocalyptic framing may be used to shape regulation or public perception in ways that benefit specific companies or agendas.

The Guardian's examination presents six separate claims and the reactions they have provoked, without settling on a single verdict about the overall likelihood or severity of AI-related catastrophe. The newspaper's report indicates that the debate remains unresolved among specialists, with sharply differing views on how much weight the public and policymakers should give to warnings of existential risk from artificial intelligence.

The Guardian did not attribute the individual claims to named sources in the excerpt reviewed, and further detail on the identities and arguments of the six experts referenced was not immediately available.