Will Cruz’ One Weird AI/BEAD Trick Work?
What’s New: Senator Cruz has gained Senate Commerce Committee support for a proposal to tie a state’s eligibility for BEAD funding to the state agreeing not to pass AI related legislation for ten years. In this note we analyze the prospects for the proposal as well as how the recent changes Commerce has imposed on the BEAD program undercut Cruz’ efforts.
Background. A popular internet meme involves “one weird trick” to enable one to do things (lose weight, make millions, grow hair) that previously seemed undoable. DC has its own version a one weird trick that enable government officials to do things despite logic, laws, or other gating factors. We pointed to one last week enabling the Secretary of Commerce to transfer funds from fiber providers to satellite services.[1] This week we focus on another weird trick that affects both AI development and BEAD. The background is as follows.
The House passed a State AI legislation moratorium.
The House version of the reconciliation bill included a provision to ban State legislation regulating AI for ten years.[2]
But that provision may violate the Byrd Rule.
The Byrd Rule prohibits the inclusion of non-budget “extraneous matters” in a reconciliation bill.[3]
Cruz proposes an alternative that might get by the Byrd Rule.
In an effort to avoid the Byrd rule, Cruz proposed that no BEAD funds could flow to states that do not agree to the ten-year moratorium.[4]
The bill also provides $500 million for BEAD, adding language that ties the broadband deployment to AI.[5]
The bill has passed through the Senate Commerce Committee.
The provision will be sent to the parliamentarian soon for a Byrd rule review.[6]
Analysis
There is a good chance that the Parliamentarian rules that the Cruz Amendment does not comply with the Byrd Rule.
We are not experts in the Byrd rule, but we think it more likely than not that the Parliamentarian views the Cruz Amendment as extraneous to the budget.
While the tie to BEAD and the $500 million have budget implications, the Parliamentarian may see the language as primarily designed to achieve a policy outcome and not a budget outcome.
Still, the Cruz amendment is much more likely to gain Parliamentarian approval than the House version, which was highly likely to be struck out of the bill.
There is a good chance that there will be enough Republican opposition to the Cruz Amendment that it does not get included in the Senate Bill.
Republican Senator Hawley has already announced opposition to the Cruz Amendment, with Republican Marsha Blackburn also indicating concerns.
Other Republicans, particularly former Governors and others who prioritize states’ rights, may join in opposition.
Most, if not all, the Democrats will also oppose the amendment.
Assuming the Cruz amendment survives both the “Byrd bath” and the opposition in the Senate, the question is what will be the impact in the states.
The general view is that states will care more about connecting rural communities than passing AI legislation.
We are not sure that is correct.
One factor that the general view overlooks is that the new BEAD rules materially reduce the incentives for states to choose BEAD over AI legislation.
Under the new BEAD rules, most of what the states are going to receive they would receive without BEAD funding.
If one assumes that most of the BEAD funding will now go to satellite providers,[7] then states will be in a position in the BEAD funding arguably gets the rural areas broadband coverage.
But without BEAD funding, they still have satellite broadband coverage.
They only lose some subsidy for equipment and a theoretical advantage of reserve spectrum capacity.
Moreover, some key incentives for a Governor to participate in BEAD are now unlikely.
There is a positive benefit to Governors to attending broadband ground breakings. There is none for bringing something to a rural community that the rural community already had.
There is a positive budget impact for Governors of using funds not needed for deployment for other related matters. But under the new rules, those funds are likely to be returned to the US Government.
While the calculus might be different for Democrat and Republican governors, no matter the party affiliation of the Governor, the trade-off of BEAD funding versus AI legislation is now different because of the rule changes governing BEAD.
State AI legislation is generally popular.[8]
Also, telling the federal government that they should not control state policy is also generally popular.
Bottom Line:For investors in AI hoping for the state legislative moratorium, the issues related to the Byrd rule and the political forces aligned against the moratorium suggest that adoption is still an uphill climb. But in addition, the new BEAD rules undercut the leverage that the Cruz amendment sought to use to force states to drop their AI legislation. While it is difficult to have conviction about the ultimate outcome—there are too many variables—the calculus that Cruz and those urging the moratorium were using to believe that states would absolutely prefer BEAD funds to AI legislation has been directionally diminished.
[1] The rules do not qualify satellite performance for the purpose of defining an unserved area. The rules do qualify satellite performance for the purpose of qualifying for funding. That seems inconsistent, but if satellite performance qualified in determining whether an area was unserved, it could result in all rural areas being defined as served, which would not serve the political interests of the Trump Administration. On the other hand, if satellite were not qualified for funding, that, apparently, would also not serve the interests of the Trump Administration.
[2] The provision reads “no State or political subdivision thereof may enforce, during the 10-year period beginning on the date of the enactment of this Act, any law or regulation of that State or a political subdivision thereof limiting, restricting, or otherwise regulating artificial intelligence models, artificial intelligence systems, or automated decision systems entered into interstate commerce.”
[3] A summary of the Byrd rule can be found here.
[4] The relevant language is “RECEIPT OF FUNDS CONDITIONED ON COMPLIANCE WITH MORATORIUM.—On and after the date of enactment of this subsection, no amounts made available to carry out this section may be obligated to an eligible entity or a political subdivision thereof that is not in compliance with subsection (q). “ Section (q) is as follows: MORATORIUM.— ‘‘(1) IN GENERAL.—Except as provided in paragraph (2), no eligible entity or political subdivision thereof may enforce, during the 10-year period beginning on the date of enactment of this subsection, any law or regulation of that eligible entity or a political subdivision thereof limiting, restricting, or otherwise regulating artificial intelligence models, artificial intelligence systems, or automated decision systems entered into interstate commerce.” There are some limited exceptions such as laws that remove legal obstacles to deploying AI. In addition, the language about the $500 million reads that the funds can be used for “the construction and deployment of infrastructure for the provision of artificial intelligence models, artificial intelligence systems, or automated decision systems.”
[5] It is not clear how this money will be used but we find it amusing that it was included. Looking at the language, we think it is not serious as it is a small amount of money designed to accomplish a large amount for which the states have no model, no background, no institutional support and no plan. But perhaps the bigger point is that as a result of Commerce’s decision to adjust the rules to favor satellite and the implication that it will require all non-deployment funds to be returned to the federal government, billions of the $42.5 billion of BEAD funds will not be utilized by the states. In that light, the money offered by Cruz will be irrelevant in the real world, other than for making the argument that the amendment is budget related.
[6] This is a process known as a “Byrd bath.” As far as we know, it is the only DC insider phrase that is not an acronym and therefore we find it lovely by comparison.
[7] It is difficult to have a high level of confidence in precise numbers but noted broadband expert Doug Dawson wrote that “We can only guess how many fiber grant requests that will kill – but it’s not hard to imagine these rules killing 80% or 90% of fiber awards.” We know lots of states are doing calculations now, some of which agree with Mr. Dawson and some of come to different conclusions.
[8] A recent poll found that “Voters — including Republicans — do not want the budget reconciliation bill to pre-empt state efforts on AI: (i) 81% of voters overall agree that "advances in AI are exciting but also bring risks, and in such fast-moving times, we shouldn't force states to sit on the sidelines for a full decade." This includes 78% of Republicans; (ii) 73% of voters want both states and the federal government to regulate AI; (iii) Most voters (59%) oppose including a provision into the budget bill that would preempt states from being able to regulate AI for a ten-year period. By a 19-point margin, more Republicans oppose (50%) than support (31%) this provision as well; and (iv) A majority of voters (52%) say the provision makes them less likely to support the budget bill, with only 15% of voters saying the provision makes them more likely to support the budget bill.