A federal proposal that would ban states and local governments from regulating AI for five years could soon be signed into law, as Sen. Ted Cruz (R-TX) and other lawmakers work to secure its inclusion into a GOP megabill — which the Senate is voting on Monday — ahead of a key July 4 deadline. 

Those in favor — including OpenAI’s Sam Altman, Anduril’s Palmer Luckey, and a16z’s Marc Andreessen — argue that a “patchwork” of AI regulation among states would stifle American innovation at a time when the race to beat China is heating up. 

Critics include most Democrats, many Republicans, Anthropic’s CEO Dario Amodei, labor groups, AI safety nonprofits, and consumer rights advocates. They warn that this provision would block states from passing laws that protect consumers from AI harms and would effectively allow powerful AI firms to operate without much oversight or accountability. 

On Friday, a group of 17 Republican governors wrote to Senate Majority Leader John Thune, who has advocated for a “light touch” approach to AI regulation, and House Speaker Mike Johnson calling for the so-called “AI moratorium” to be stripped from the budget reconciliation bill, per Axios.

The provision was squeezed into the bill, nicknamed the “Big Beautiful Bill,” in May. It was initially designed to prohibit states from “[enforcing] any law or regulation regulating [AI] models, [AI] systems, or automated decision systems” for a decade.

However, over the weekend, Cruz and Sen. Marsha Blackburn (R-TN), who has also criticized the bill, agreed to shorten the pause on state-based AI regulation to five years. The new language also attempts to exempt laws addressing child sexual abuse materials, children’s online safety, and an individual’s rights to their name, likeness, voice, and image. However, the amendment says the laws must not place an “undue or disproportionate burden” on AI systems — legal experts are unsure how this would impact state AI laws.

Such a measure could preempt state AI laws that have already passed, such as California’s AB 2013, which requires companies to reveal the data used to train AI systems, and Tennessee’s ELVIS Act, which protects musicians and creators from AI-generated impersonations. 

But the moratorium’s reach extends far beyond these examples. Public Citizen has compiled a database of AI-related laws that could be affected by the moratorium. The database reveals that many states have passed laws that overlap, which could actually make it easier for AI companies to navigate the “patchwork.” For example, Alabama, Arizona, California, Delaware, Hawaii, Indiana, Montana, and Texas have criminalized or created civil liability for distributing deceptive AI-generated media meant to influence elections. 

The AI moratorium also threatens several noteworthy AI safety bills awaiting signature, including New York’s RAISE Act, which would require large AI labs nationwide to publish thorough safety reports.

Getting the moratorium into a budget bill has required some creative maneuvering. Because provisions in a budget bill must have a direct fiscal impact, Cruz revised the proposal in June to make compliance with the AI moratorium a condition for states to receive funds from the $42 billion Broadband Equity Access and Deployment (BEAD) program.

Cruz released another revision last week, which he says ties the requirement only to the new $500 million in BEAD funding included in the bill — a separate, additional pot of money. However, close examination of the revised text finds the language also threatens to pull already obligated broadband funding from states that don’t comply.

Sen. Maria Cantwell (D-WA) previously criticized Cruz’s reconciliation language, claiming the provision “forces states receiving BEAD funding to choose between expanding broadband or protecting consumers from AI harms for ten years.”

What’s next?

As of Monday, the Senate is engaged in a vote-a-rama — a series of rapid votes on the budget bill’s full slate of amendments. The new language that Cruz and Blackburn agreed on will be included in a broader amendment, one that Republicans are expected to pass on a party line vote. Senators will also likely vote on a Democrat-backed amendment to strip the entire section, sources familiar with the matter told TechCrunch.

Chris Lehane, chief global affairs officer at OpenAI, said in a LinkedIn post that the “current patchwork approach to regulating AI isn’t working and will continue to worsen if we stay on this path.” He said this would have “serious implications” for the U.S. as it races to establish AI dominance over China. 

“While not someone I’d typically quote, Vladimir Putin has said that whoever prevails will determine the direction of the world going forward,” Lehane wrote. 

OpenAI CEO Sam Altman shared similar sentiments last week during a live recording of the tech podcast Hard Fork. He said while he believes some adaptive regulation that addresses the biggest existential risks of AI would be good, “a patchwork across the states would probably be a real mess and very difficult to offer services under.” 

Altman also questioned whether policymakers were equipped to handle regulating AI when the technology moves so quickly. 

“I worry that if … we kick off a three-year process to write something that’s very detailed and covers a lot of cases, the technology will just move very quickly,” he said. 

But a closer look at existing state laws tells a different story. Most state AI laws that exist today aren’t far-reaching; they focus on protecting consumers and individuals from specific harms, like deepfakes, fraud, discrimination, and privacy violations. They target the use of AI in contexts like hiring, housing, credit, healthcare, and elections, and include disclosure requirements and algorithmic bias safeguards.

TechCrunch has asked Lehane and other members of OpenAI’s team if they could name any current state laws that have hindered the tech giant’s ability to progress its technology and release new models. We also asked why navigating different state laws would be considered too complex, given OpenAI’s progress on technologies that may automate a wide range of white-collar jobs in the coming years. 

TechCrunch asked similar questions of Meta, Google, Amazon, and Apple, but has not received any answers. 

The case against preemption

“The patchwork argument is something that we have heard since the beginning of consumer advocacy time,” Emily Peterson-Cassin, corporate power director at internet activist group Demand Progress, told TechCrunch. “But the fact is that companies comply with different state regulations all the time. The most powerful companies in the world? Yes. Yes, you can.”

Opponents and cynics alike say the AI moratorium isn’t about innovation — it’s about sidestepping oversight. While many states have passed regulation around AI, Congress, which moves notoriously slowly, has passed zero laws regulating AI.

“If the federal government wants to pass strong AI safety legislation, and then preempt the states’ ability to do that, I’d be the first to be very excited about that,” said Nathan Calvin, VP of state affairs at the nonprofit Encode — which has sponsored several state AI safety bills — in an interview. “Instead, [the AI moratorium] takes away all leverage, and any ability, to force AI companies to come to the negotiating table.”

One of the loudest critics of the proposal is Anthropic CEO Dario Amodei. In an opinion piece for The New York Times, Amodei said “a 10-year moratorium is far too blunt an instrument.” 

“AI is advancing too head-spinningly fast,” he wrote. “I believe that these systems could change the world, fundamentally, within two years; in 10 years, all bets are off. Without a clear plan for a federal response, a moratorium would give us the worst of both worlds — no ability for states to act, and no national policy as a backstop.”

He argued that instead of prescribing how companies should release their products, the government should work with AI companies to create a transparency standard for how companies share information about their practices and model capabilities. 

The opposition isn’t limited to Democrats. There’s been notable opposition to the AI moratorium from Republicans who argue the provision stomps on the GOP’s traditional support for states’ rights, even though it was crafted by prominent Republicans like Cruz and Rep. Jay Obernolte.

These Republican critics include Sen. Josh Hawley (R-MO), who is concerned about states’ rights and is working with Democrats to strip it from the bill. Blackburn also criticized the provision, arguing that states need to protect their citizens and creative industries from AI harms. Rep. Marjorie Taylor Greene (R-GA) even went so far as to say she would oppose the entire budget if the moratorium remains. 

What do Americans want?

Republicans like Cruz and Senate Majority Leader John Thune say they want a “light touch” approach to AI governance. Cruz also said in a statement that “every American deserves a voice in shaping” the future. 

However, a recent Pew Research survey found that most Americans seem to want more regulation around AI. The survey found that about 60% of U.S. adults and 56% of AI experts say they’re more concerned that the U.S. government won’t go far enough in regulating AI than they are that the government will go too far. Americans also largely aren’t confident that the government will regulate AI effectively, and they are skeptical of industry efforts around responsible AI.

This article was updated June 30 to reflect amendments to the bill, new reporting on the Senate’s timeline to vote on the bill, and fresh Republican opposition to the AI moratorium.

Navin Chaddha, managing director of the 55-year-old Silicon Valley venture firm Mayfield, is betting big on AI’s ability to transform people-heavy industries like consulting, law, and accounting. The veteran investor, whose wins include Lyft, Poshmark, and HashiCorp, recently discussed at TechCrunch’s StrictlyVC evening in Menlo Park why he believes “AI teammates” can create software-like margins in traditionally labor-intensive sectors, and why startups should right now target neglected markets rather than compete head-to-head with giants like Accenture — though he acknowledged that disrupting outfits where relationships and trust matter is sometimes harder than Silicon Valley anticipates. This conversation has been edited lightly for length and clarity.

You think that law firms, consulting companies, and accounting services – collectively a $5 trillion market – will be completely reimagined by AI-first companies that operate with software-like margins. Prove it. What have you seen beyond PowerPoint presentations?

I think an advantage of a firm that has been in business for over 50 years is that it has seen all the trends, from mainframe to minicomputers to PCs, to the internet, to mobile, cloud, social and now this AI era. The example I would give is in the late ’90s, this concept of e-business came,  which was: if I’m a physical business, I cannot survive if I’m just brick and mortar; I need to be click and mortar. Then outsourcing became a trend, and offshoring became a big trend. You couldn’t build a software services company without a presence in India or one of the emerging markets. The same thing happened with supply chains and manufacturing — China and Taiwan rose. So what is this new era with AI? Clearly, AI is a 100x force, and AI is teaming up with humans, hopefully to make them better. And I think it is, and it’s going to help reimagine business.

A lot of the repetitive tasks are going to be done by AI… and there’ll be two models. One is that you grow organically. The second is that you grow inorganically. . . 

Can you give a specific example of how this will work?

What are the kinds of things an LLM or AI can do?  Well, say I have to implement Salesforce. Who wants to go do that work? The human will come in and say, ‘I’m your client manager. You have to implement Salesforce.’ It’s the same set of things. Use AI as the horse to do it, and whatever AI can’t do, have the human in the loop. 

Now, suddenly, if you start doing these kinds of things, you can have less work done by humans and more work done by AI, and [customers] only pay for AI when [they]  use it. 

And the market [entry] should not be to go after [big consulting and IT companies] like Accenture, Infosys, or TCS. Go after the neglected masses. There are 30 million small companies in the U.S., and 100 million worldwide that can’t afford knowledge workers. Provide them service as software. They say, “I need a receptionist. I need a scheduler. I need somebody to build my website…” AI should be used to [create] startup funding forms, with some human [involvement] for negotiation. You don’t compete with the Accentures of the world. You go after fragmented markets, where instead of charging per hour, instead of charging per month for a contractor, you charge per event.

So outcome-based pricing rather than time-based billing.

This is outcome based, yes . . . Cloud billing is like that; electricity is like that . . .If 80% of the work will be done by AI, it can have an 80% to 90% gross margin. Humans can still have a 30% to 40% margin. You could have blended margins of 60% to 70% and produce 20% to 30% net income. And believe me, most services companies make money. Tech companies don’t. They live on venture money and then public market money. 

You just led the Series A for a company called Gruve a few weeks ago. It’s an AI tech consulting startup. What did you see in its early customer pilots?

I think this is where the combination of inorganic and organic happens. [Gruve was founded by] very successful founders who had done two services companies before [and] bootstrapped, and got them to $500 million in revenue each, and $50 to $100 million in profits. They started this time and said, ‘What do we know? We know security.” So they acquired a $5 million security consulting company [that offers managed security services]. And they said, “Let’s look at the people. All the growth from this point on will happen through AI.” And they grew that from [$5 million in revenue] to $15 [million in revenue] in six months. They literally have an 80% gross margin. It’s outcome-based. Customers love it. Cisco loves it. They say, “Hey, I’m not getting hacked. Why am I paying for all these security people?” If you outsource, [a vendor has traditionally charged] $10,000 a month. [Gruve] says, “ [You pay us] zero. If you get hacked, if there is an event, if I look at it, then you pay me.”

Can’t companies like McKinsey just buy these AI capabilities? They’ve got big businesses they don’t want to lose.

Yeah, I think what’s going to happen is this is where the innovator’s dilemma comes in. When enterprise software companies, which were perpetual license companies, saw SaaS companies emerging, they didn’t want to adopt [the model] because [SaaS companies] charge companies monthly instead of five years up front. The enterprise companies also collected a 20% maintenance fee. It was hard [for them] to get off that drug and to say, “Oh, I’ll charge you monthly.” The business model innovation was the key thing. They didn’t do it. So McKinsey and Accenture, with so much dislocation, they’re going to be busy serving their clients [which is why I advise founders to] go after the neglected masses. Figure out a unique go-to-market strategy and service somebody they [an Accenture can’t come down market to serve].

But they’re going to get reimagined too. So these small companies, which are not competing with them today, mark my words: in 10 years, they will be competing with them. And those big companies – McKinsey, BCG, Accenture, TCS, Infosys – all have the innovator’s dilemma [and are asking themselves]: when do I do it? [When do I switch to an outcome-based AI model?] Because as a public company, my revenues are going to go down from predictable revenue to utility-based revenue.

You carved out $100 million from your recently raised funds to dedicate to “AI teammates” last fall. What makes a true AI teammate versus an AI tool?

There’s a lot of buzzwords in the industry. First it was copilots, then AI tools, AI agents, AI teammates. So the Mayfield thesis is that an AI teammate is a digital companion that collaborates with a human on shared goals and gets to better outcomes. The technology it might be built on could be agentic technologies or copilots. The manifestation of it is, “I’m an HR teammate. I’m a sales engineering teammate.” The aim is not to replace; the aim is to team up and collaborate together.

When people started talking about teammates and assistants, it sounded novel, but I wonder if that’s going to look callous as more people lose their jobs. Does Silicon Valley have a marketing problem?

Absolutely right, and I think we need to not sugarcoat it. We need to address it head-on. . .Yes, there’s going to be job displacement, but humans are smart. They’re the jockey. The horse here is AI. We will reimagine ourselves. We will reinvent ourselves. Today, the focus is on cutting costs, but we will figure out how to expand our markets, how to increase revenue. This happens with every technology wave that comes. When Microsoft Word came to PCs on the desktop, people thought [executive assistants] were out of business. Then Excel came, and accountants who did calculations — everyone thought they were out of business. We saw the same with Uber and Lyft. People thought taxi drivers would go away. But what happened instead? The markets expanded.

My thesis is, the way emerging markets like India, China, and Africa never had landlines — you couldn’t dig copper, so they went wireless, cellular — that’s what’s going to happen with many markets. AI will do the work where humans are not even available to serve that customer. So, long-run, I’m very, very bullish. In the short-run, there will be pain, but no pain, no gain.

Speaking of coding, a recently announced “vibe-coding” deal centered on a six-month-old Israeli company that had just reached 250,000 users per month and $200,000 in monthly revenue. It was bought by another Israeli company, Wix, for $80 million in cash. Does that math make sense to you?

Actually, these days, no math makes sense. We’re in the AI age. You don’t know what’s going to happen. I’m surprised that with $2.4 million in [annual recurring] revenue they only sold for $80 million. I thought it would be $800 million, right? [Laughs.] In today’s world, you don’t know. It’s a marketplace.

How do you invest in that market?

That’s where the secret recipe comes from people who are proven investors. They’ve cracked the code. It’s not a science; it’s an art. It’s like the 10,000-hours [rule]: the more you practice this, the better you get. And the firms that have been around for 50 or 60 years – we’ve seen all kinds of bubbles. 

The number-one rule is, have your own North Star. Have discipline and have no FOMO, because FOMO is for sheep. And if you have those two or three things, your own strategy and no fear, [you’ll do well]. Just remember one thing: for people [in this audience] who are VCs, we’re in the money management business. We’re not about collecting logos. We are about taking small amounts of money and making them bigger. 

During this part [of the cycle], a lot of money will get made. But I think 80% of the people are going to lose money. They don’t know what they’re doing.