Reed Hundt: An Artificial Intelligence-Focused Telecom Act
Congress will likely pass AI and data center legislation, but recent Supreme Court rulings threaten the bipartisan agency structure needed to implement it well.
I think there is a near certainty that Congress will enact a law regarding artificial intelligence labs and a high likelihood that will pass a bill concerning data centers.
I don’t know when these statutes will emerge. It is probable that the next Congress and not this one will vote for these measures.
Why AI legislation Is coming
The reasons for this prediction are:
Just as the dawn of both the commercial Internet and digital cellular called for forward looking government shaping of markets for competitors, new entrants, customers and suppliers, so too AI – which is the Internet to the 10th power more or less—gives rise to the same need – although the desirable statute would be quite bit different than the 1992 Cable Act or the 1993 auction law or the 1996 Telcom Act.
It is in the interest of the labs – especially Open AI and Anthropic – to have predictability concerning government action. Ad hoc decisions by government threaten the ability of businesses to plan and given the tens of billions of dollars that the labs are investing those plans matter not just to them but to the whole stock market and the economy. A law resembling the 1996 Telcom Act in scope and mandate would benefit the labs.
The labs need the United States government to make sure that their work product is not pilfered by bad actors. The Chinese firms may be bad acting right now. A law is a way to define the relative roles of the labs and the United States government.
The labs need data centers. They have to be built and rebuilt and renovated. We need maybe 10,000 of them in the United States. We don’t want them in other countries instead of here. But they are very unpopular with voters, because they consume land, transportation, water, and electricity, driving up prices to consumers and other businesses. That’s why some federal and state officials are calling for moratoriums. Moratoriums are horrible for the labs and the data center developers.
The case for a national data center auction
The labs and the data center developers and all electricity consumers would benefit from a law that enabled the national government to conduct a data center location auction. In that auction the government would make a match between local government and data center developers. In this national auction the government would create contracts between the developers that offered benefits to the community (especially new cheap clean power) and the communities that will give the developers the land, zoning, and infrastructure that they need.
This is the same type of two-sided auction used to match broadcast tv license sellers with wireless communication buyers, with the FCC as the intermediary.
Communities that don’t want centers need not bid. But they won’t get the benefits of cheap clean power. Developers can count on being able to build and get power, in return for benefiting the local community.
This is better than a moratorium. It is better than telling developers to provide power only to themselves. And it is better for the country, since we really do not want Chinese firms to dominate this brave new world.
The labs also need an expert agency to work with. It should report to Congress, the President and the American people. It should be staffed with people who deeply understand this amazingly complex technology. According to Claude (who should know): Under the 2023 framework, a single executive order assigned AI responsibilities to more than 50 agencies; the current framework disperses those responsibilities across several narrower orders, with about a dozen agencies holding named roles and procurement/grant provisions still reaching government wide.
And more than a dozen Congressional committees have an oversight role of some kind.
The technical complexity of AI is matched only by the complex effects on business, society and humanity that the technology itself portends to have. It would really be a good idea for at least one single body of public officials to work closely with AI firms to understand, at the very least, what is happening and what could be happening.
The Supreme Court's blow to bipartisan agencies
Lastly, I want to comment on the Supreme Court.
In the Slaughter case the Court decided that the President can kick any commissioner out of a multimember agency. The meaning of this is that the Court has discouraged bipartisan decision-making.
Many commentators have decried the decision as the end of independent agencies. It is indeed surprising that the clear Congressional intentions of more than a century are so blithely rejected by the Court with hardly a smidgeon of rationale. But it’s the loss of bipartisanship and not independence that is the major injury on good government the Court has inflicted.
The primary purpose of multimember commissions is to apply the intent of Congress expressed in legislation to the situations that new facts constantly create. This is done mostly by enacting rules and partly by enforcing them. In this critical activity – where the rubber of law meets the road of reality – a bipartisan Commission reenacts the bipartisan process that caused the law itself to be passed.
The President and the rest of the Executive Branch isn’t involved in most cases. The President, after all, signed the bill into law (except when like the 92 Cable Act it was passed over a veto) and trusts the chosen chair to implement it consistent with the President’s overall policy direction and, crucially, the actual language of the law.
But that process is bipartisan because the Commission, in its mini reenactment of Congress, votes on the rules.
Because the commissioners from the opposing party have lots of ways to give the Chair a hard time in rulemaking, virtually all chairs most of the time must and do try to obtain bipartisan consensus.
As FCC chair I worked hard to have almost 100 percent of our decisions be unanimous. It was partly to get things done quickly at the Commission. It was partly because the Republicans controlled Congress after the 1994 election and I wanted to get along with them. It was partly because dissension among us would open widely the door to judicial reversal of our decisions. And it was partly because we were more likely to obtain business concurrence with our general direction if I could show a united front among Republican and Democratic commissioners.
To get this result required lots of persuasion and compromise. We were bipartisan in practice.
Oddly, while bolstering the likelihood of agencies to rule in a partisan fashion, in other rulings the Court has undercut the ability of agencies, even when acting exclusively and partisanly at the direction of the president, to address emergent policy and administrative issues.
The agencies’ decisions are not to be respected by the judiciary nearly as much in the past.
The agencies are limited in their capacity to interpret Congressional mandates.
The invaluable Senior Executive Staff may not be able to retain their positions.
In these other, earlier rulings, the Court has made it hard even for a unitary executive to run an effective government.
What then is the Court trying to accomplish? It works against bipartisanship (in a mirror of its own behavior as a court). It undercuts the outcomes of agency decisions even as they are more than ever before an expression of the President’s wishes in interpreting statutes. Yet there is no doubt that without rules no laws can become effective.
In conclusion, I cannot discern the structure and function of the Executive Branch that the Court wishes to impose. I do wonder about the quality of its collective experience when it acts as a management consulting firm.
Reed Hundt is a former chairman of the Federal Communications Commission, where he served from 1993 to 1997 under President Bill Clinton and Vice President Al Gore. During his tenure he oversaw implementation of the Telecommunications Act of 1996, championed spectrum auctions, and launched the E-Rate program to connect schools and libraries to the internet. A graduate of Yale University and Yale Law School, he practiced law before joining the FCC. After leaving the commission he became an author and venture investor, and later co-founded and led the Coalition for Green Capital, an organization focused on clean energy financing through green banks. He has written several books on economic and telecommunications policy. He is CEO of the non-profit group, ReFounding America. This Expert Opinion is exclusive to Broadband Breakfast.
Broadband Breakfast accepts commentary from informed observers of the broadband scene. Please send pieces to commentary@breakfast.media. The views expressed in Expert Opinion pieces do not necessarily reflect the views of Broadband Breakfast and Breakfast Media LLC.
