Every few weeks, we hear from the AI community about new features that extend thinking. With excitement and fear, we try to speculate what this means for humanity and society. Predict which jobs will disappear, which ones will appear, and when computers will surpass human reason. These topics have become part of our daily conversation.
What has received little attention, but is fundamentally important, is the simultaneous announcement of new regulatory measures by numerous governments. More than 1,500 bills are being considered in each state. There are hundreds of members in Congress, dozens of executive actions in the executive branch, and market participants have diverse views on how the AI industry should be regulated.
I once thought that this AI revolution would be controlled by the invisible hand of the market. These approximately 2,000 proposals show that this is not possible.
Many of these proposed policy changes are thoughtful, necessary, and good steps forward, but collectively they are not enough. The commonality is that they all focus on some of today’s issues. None of the proposals seeks to establish a permanent, comprehensive and forward-looking regulatory framework.
National regulatory bodies are traditionally established after a crisis. The SEC was created after the 1929 crash, and the Nuclear Regulatory Commission (NRC) was created after the partial meltdown at Three Mile Island.
Don’t wait for a crisis to occur during this AI revolution.
We need policymakers to step forward and create national regulatory bodies with broad powers to adequately oversee the AI revolution for the opportunities and challenges of today and tomorrow.
The establishment of a national regulatory body is not a panacea. Some would argue that the so-called AIR Commission has the most dynamic mandate, as AI will impact every sector of society. Regulatory authorities tend to overregulate, and continued parliamentary, executive, judicial, and public oversight is important. The AI revolution is a global competition, and balancing innovation-driven markets with appropriate regulation is a tough task we must master.
However, as a partial step forward, I would like to suggest one aspect of the SEC’s operating model.
Prior to joining Nasdaq, he ran an entrepreneurial software company. Speed to market is of the utmost importance and we are determined to update our products as quickly as possible. I was shocked to learn that in order to implement improvements to its core exchange technology, Nasdaq is required by SEC rules to submit details of these changes to the SEC, which will make them public for comment and subsequent review. The fact that our competitors knew exactly what was in the next release was totally off-putting.
Over time, this method of operation resulted in the fact that the US capital market is the best in the world. SpaceX’s IPO was only possible in the US market.
I am confident that the public comments received regarding significant changes to the LLM model will be orders of magnitude smaller than the comments received regarding changes to exchange order types.
In advocating for national regulation of AI, we don’t know whether we’re claiming this is a great good or a minor evil, but we’re sure this is just the beginning. As the AI revolution progresses, regulatory mechanisms will also need to change. The SEC of 2026 is a faint echo of the one created in 1934.
I believe that in the short term any regulatory efforts will impede progress, but in the long term good rules of the road will create the conditions for greater good.
—Bob Greifeld, Managing Director and Co-Founder of Cornerstone Financial Technology, CNBC Contributor
