Earlier this year, the National Institute of Standards and Technology (NIST), through it’s U.S. Artificial Intelligence Safety Institute, wrote a draft paper on the potential “misuse” of certain AI models. NIST then requested comments from the public and relevant stakeholders regarding their paper. While the paper was required of them through President Biden’s AI executive order, the tone and specific content were not. The draft product is unfortunately pessimistic and fails to adequately contend with the many potential upsides to AI adoption.
In filing a comment in response to the draft paper, NetChoice has focused on:
- The dangers of setting a pessimistic tone for AI;
- The reality that AI is already aggressively regulated in the United States;
- The need to improve legal definitions of AI in order to avoid over-regulation; and
- The fact that NIST does not cite appropriate authority to promulgate new AI-specific regulation.
Our government should be focused on winning the AI race against our foreign adversaries, not on scare tactics and red tape. In the instances where there are genuine, novel harms related to AI development, Congress must act—not executive branch agencies.