- Congress’ new budget bill paves the way for more state AI regulation
- Analysts say AI regs will probably have minimal impact on telcos – for now
- But various state laws could complicate the adoption of national AI regulation
The “One Big Beautiful Bill” is officially legal without a proposed 10-year ban on states regulating AI. While more AI laws may be on the horizon, experts think that probably won’t impact telecom companies using AI – at least not yet.
The idea behind AI regulation is to ensure the AI (specifically, agentic AI) doesn’t do anything harmful, said J.Gold Associates Founder Jack Gold, “like the reports of LLMs and GenAI systems trying to threaten their operators if they think they are being shut down.”
Telcos, who are still figuring out how to monetize and effectively use AI, simply provide the network platform AI applications ride on. “So, any regulations would fall on the companies deploying the AI,” Gold explained.
If there is a regulatory risk to telcos, it’d be related to privacy and security. Operators “do gather a ton of information on customers, and any exposure, as has already happened, is problematic,” he noted.
“Could AI make that even worse? Possibly…,” said Gold.
Legislators across the board were no fans of the AI law moratorium, as the U.S. Senate last week voted 99-1 to scrub the provision from the big budget bill.
The Senate also sought to tie the AI moratorium to Broadband Equity, Access and Deployment (BEAD) funding. If the provision had remained, states could have been forced to pick and choose between enacting AI laws or getting their BEAD allocations.
State AI proposals
State and local AI oversight has ramped up in the last couple of years, as almost all 50 states have proposed some kind of AI regulatory framework. At least 17 states have already enacted AI regulations, said New Street Research Policy Analyst Blair Levin.
California for instance enacted a law requiring companies that create or distribute GenAI systems to disclose when content is AI-generated, as well as provide tools for users to detect that content.
Colorado and Utah have passed similar legislation, with Utah requiring certain businesses (like those that collect financial or health data) to disclose when a consumer is interacting with AI and not a human.
“I don't think any of those laws have had a material effect on telecom enterprises,” Levin told Fierce. He noted however AI will eventually “increase the value of the networks over which AI applications must ride.”
As AI technology advances and GenAI-enabled threats are more “broadly realized,” regulations must evolve as well, said neXtCurve Executive Analyst Leonard Lee. Meaning telcos could see a bigger impact from AI regs in the years to come.
“The telecom industry will be at the epicenter of this evolution,” he said. “Agentic and AI-augmented spam marketing and the rights to be forgotten and to be deleted will likely be hot topics that can have broad and profound impact on an operator’s AI business ambitions.”
But states passing different types of AI laws makes “national scaling of AI-related services more complicated,” Lee added, due to states having their own compliance requirements and rules for privacy and copyright protection.
Similarly, Levin said he thinks federal legislation concerning AI regs would have been better. However, “the states are justified in believing that as long as Congress is incapable of passing anything important, other than tax and budget legislation, states should be free to act to protect their residents,” he concluded.