Before regulating AI, define strict limits for the government

The piece identifies three AI-regulatory approaches that it says are gaining traction. These are Sam Altman's IAEA-style international body, a House proposal for government-licensed verification organizations, and Google's industry-funded self-regulatory proposal. The op-ed's title is "Before regulating AI, define strict limits for the government."

The op-ed describes the three approaches as contenders for how AI should be governed, but does not endorse any of them. Instead, the author contends that Congress needs to check agencies' powers before such bodies are invented. The argument is that new regulatory structures should have clearly defined boundaries set in advance, rather than being created with open-ended authority. The author does not name a specific agency now in existence, but refers to agencies that could be created by these proposals.

The piece treats the current moment as a window in which limits can still be set before any new regulatory body exists. Its central recommendation is that Congress act preemptively to constrain future agencies. The op-ed does not provide a detailed list of what specific limits it wants, but frames the issue as one of timing and institutional design. It does not compare the three proposals in depth, nor does it endorse Altman, the House, or Google as the preferred route. Instead, it argues that the design of any new regulator must be constrained by prior congressional action.