Trump Says AI Risks Can Be Managed and He Would Work With Democrats on Guardrails

Asked whether he would work with Democrats on AI safety, Trump replied, "I would absolutely", according to the CBS transcript. "I don't mind guardrails. But I don't wanna put us in a deep second position. We're leading number one," he said. "And China has no guardrails, by the way."

The comments are the first time since the midterm elections that Trump has signalled a willingness to negotiate on AI rules with the opposition, which will have more leverage after its gains at the polls. They leave open what kind of safeguards he might accept. Trump has generally favoured letting technology companies police themselves over sweeping government restrictions, and it remains unclear at what level he would back any new rules, Reuters reported.

The interviewer, Ross Douthat, put to Trump that people in Silicon Valley believe his decisions on AI will be what he is remembered for, and asked why he was not worried by the warnings of AI executives who say their own technology could pose a risk to humanity. Trump framed his answer around competition with Beijing. "Because China's gonna have it, okay? So if we don't have it, China's gonna have it," he said. "That would put us at an unbelievable disadvantage. It's all between us and China. There's nobody else."

Pressed on whether the race with China justified the risk or whether he thought the risk was not real, Trump said: "I think that the risk is not nearly as real as it-- as they talk. And I think the risk can be managed."

Trump described himself as an "AI bull" and said one reason was his recent meeting with Chinese President Xi Jinping, who he said had left "two weeks ago" and was "all in" on what Trump called "SI", or superintelligence. "I've said whoever wins that race wins," Trump said. When Douthat noted that China's restraints come from being a totalitarian state, Trump said Beijing has "the guardrails that whatever President Xi says goes", adding: "With me, it's pretty strong, but it's not quite that strong."

Trump's stance on safeguards has shifted before. In September he called concerns about AI a "hoax" and rejected the need for guardrails, NBC News reported, before telling reporters days later that guardrails could be put up, according to The Washington Times.

The interview gives no detail on which safeguards Trump might accept, whether federal rules would override state AI laws, or how any deal with Democrats would treat the data-centre build-out that drew opposition from some voters this year. Those questions are likely to surface when the new Congress begins setting its technology agenda.