Trump was asked in an interview this week whether the United States should accept the risks posed by AI in order to stay ahead of China. His answer was: the risks are controllable, "I don't mind setting up regulatory guardrails," and he also expressed his willingness to work with the Democratic Party to establish safeguards.
This is different from his previous stance. Previously, the White House basically asked technology companies to restrain themselves and opposed comprehensive restrictions at the federal level. The premise remains the same: He doesn’t want regulation to turn the United States into a “far second place.”
On the other hand, the U.S. Department of Justice announced that the "Superintelligence Task Force" (SINTF) has begun operations. This working group was authorized by Trump in a memorandum on October 13 and is led by the Director of National Intelligence. The task force is very straightforward: to maintain the United States’ leadership in AI and counter foreign “abuse” of AI.
It would coordinate all of the Justice Department’s criminal and civil cases involving AI, while also participating in legislative proposals to block so-called “burdensome regulation.” Details of AI regulation will be announced later this month, Defense Department CTO Emil Michael said.
Looking at the two things together, the United States' thinking is becoming more and more like "loose internally and tight externally": it is willing to talk about some guardrails with local companies, but the main line is still competition and security. After the midterm elections, the Democrats won both houses of Congress, and there will only be more pushback on AI in Congress.