As AI incidents like the Hugging Face breach mount, Trump is holding to a light-touch approach to regulation while tech leaders like Dario Amodei push for stronger safeguards.
Artificial intelligence is increasingly becoming a political issue, especially as governments try to decide how much control they should have over rapidly developing AI systems. President Donald Trump has generally pushed for a light-touch approach to AI regulation, arguing that excessive rules could slow American technological development. At the same time, recent AI incidents have increased pressure from some technology executives and government officials for additional oversight.
Trump said on September 21 that there are ways to "rein" in AI if necessary, while still emphasizing his preference for limited regulation. His position comes as several major technology leaders have called for greater government involvement. Anthropic CEO Dario Amodei has pushed for stronger safeguards, while OpenAI CEO Sam Altman and Elon Musk have also discussed the need for government oversight in different contexts.
The debate became even more complicated after incidents such as the Hugging Face breach and other cases involving AI agents. Treasury Secretary Scott Bessent recently argued that AI developers should be held accountable for what their systems do and pointed to the Hugging Face incident as an example of why developers cannot simply avoid responsibility when their models cause problems. The United States and China have also agreed to continue discussions about AI safety, including protocols for dealing with incidents involving autonomous AI systems.
What makes the situation interesting is that there is not simply a debate between people who like AI and people who dislike it. The disagreement is increasingly about how quickly AI should develop and who should be responsible when something goes wrong. Trump and his administration have emphasized keeping regulation limited, while some technology executives and officials are calling for stronger safeguards. As AI becomes more capable of operating independently, that question is likely to become harder for governments to avoid.
OpenAI's own agents reportedly breached Hugging Face and uploaded malicious packages to RubyGems without direct human instruction. Here's why that's a different kind of security problem.
Read moreApple's first foldable iPhone hides its crease better than I expected, and its under-display camera is a bigger technological leap than Apple usually makes this early.
Read more