The challenges and dire threats surrounding the unchecked growth of artificial intelligence appear unlikely to move Congress or President Donald Trump to act as the U.S. Senate enters its last week before the midterms.
The rapid emergence of the technology has sparked a host of policy debates, with some warning of an existential threat to humanity, requiring urgent attention. But the only action Congress could plausibly take before November’s elections is a relatively toothless measure aiming to control the massive consumer electricity costs of AI.
“A lot of people are afraid of the dark and change is difficult,” said Sen. Jim Justice, R-W.Va. “The amazement of AI just dumbfounds us all.”
The Senate is expected to vote Wednesday on a plan to urge states to protect consumers from paying higher electricity costs from data centers. A lot of Democrats are balking at the legislation, which will need the support of 60 senators, saying the plan is weak.
Trump and House Speaker Mike Johnson are scheduled to meet with industry leaders Tuesday at the White House. Trump continues to express wariness of any new AI regulation.
Pew Research Center this summer found 71% worried AI will mean job losses over the next 20 years, up 7 percentage points from two years ago.
“Worry over job loss has shown up throughout our research about AI, including among workers and even teenagers,” a Pew analysis said.
With constituents clamoring for clarity, and midterm elections weeks away, it’s unclear where all this is eventually headed. Lawmakers are still trying to grasp the nature of the problems.
The simple reason it’s been hard to set regulations, says Darrell West, senior fellow at Washington’s Brookings Institution, is inertia. The industry has simply never dealt with significant government influence.
“They’re dealing with an industry that hasn’t been regulated for 25 years and is now very powerful,” he said. “It’s hard to regulate powerful industries.”
‘There is no one AI policy debate’
Concern about AI accelerated earlier this month when a former researcher at AI firms Open AI and Anthropic warned that the companies are “gambling with our lives.”
An Anthropic colleague then said “We really do earnestly believe AI could kill all humans!” And Anthropic’s CEO urged AI developers to slow down.
Lawmakers were quick to express concern. Sen. John Kennedy, R-La., wants a “kill switch” that would require companies to instantly shut off or curb an AI function if things look dangerous.
Sen. Bernie Sanders, an independent from Vermont, and Rep. Greg Casar, D-Texas, are seeking to pause advanced AI development until the federal government establishes a regulator to assure safe, secure progress.
A bipartisan group of senators is pushing hard for export controls, including making it tougher for China and other adversaries to get access to advanced AI.
Sens. Richard Blumenthal, D-Conn., and Josh Hawley, R-Mo., are seeking a risk evaluation program that would require developers of advanced AI to provide information to the Energy Department before deploying the technology. Hawley is chairing a subcommittee hearing this week on preventing attacks from rogue AI agents.
About all that’s getting traction is the Ratepayer Protection Act, which requires state utility commissions to consider ways to curb the electricity costs that data centers can generate.
Though it passed overwhelmingly in the House recently, it faces broad opposition from Senate Democrats as not being tough enough.
Overall, AI debate is proving to be a set of puzzles no one can solve.
“There is no one AI policy debate. There are multiple AI policy debates,” said Adam Kovacevich, founder and CEO of Chamber of Progress, a center-left trade organization representing tech firms.
He listed some of the topics: the existential risks of AI, its impact on jobs, chatbots, impacts on children, copyright, fair use.
While little is likely to happen this month, some action is likely in the future. “It wasn’t that long ago people had a positive view about technology,” said West. “Now sentiment has turned against it. The odds of something getting done are going up substantially.”
Why is Washington stuck?
The stalemate over AI stems from at least three factors:
Congress is polarized. For decades, it wasn’t unusual for Congress to take big swings to deal with emerging issues and industries.
Most recently, the Dodd-Frank banking law of 2010 provided strong new protections for consumers and helped restrain large financial institutions after the deep recession from 2007 to 2009.
In 1996, as cable television began surging, Congress passed the Telecommunications Act to promote competition in the industry.
But in recent years, such big deals have become elusive in the House and Senate.
“I have often told people as a new senator that if we were healthy as a country, if our politics were healthy, we’d be talking about AI every single day,” said Sen. Elissa Slotkin, D-Mich., in an interview.
State regulations vary from place to place.
“The worst case scenario is 50 kinds of state laws. You don’t want to write a different algorithm for different states,” said West. “Tech firms will want one set of rules.”
In Maryland last week, Gov. Wes Moore, a Democrat, set up a task force to look at new data center plans.
In Virginia, where there are more data centers than in any other state, Democratic Gov. Abigail Spanberger last week took a major step toward tight regulation of data centers, though she did not propose the moratorium many environmental groups had sought.
On the federal level, even the simplest initiatives are proving tough. Trump has insisted that efforts to control AI are a “hoax,” and has resisted any bid to regulate the industry.
“I’m not going to stifle growth of anything that’s going to be bigger than the industrial revolution,” Trump said last week. He has proposed an AI Force with a top official to lead it, but has offered no specifics.
The Ratepayer Protection Act is Congress’ current bid to get companies to pay for the electricity AI generates. While it has strong support, many lawmakers complain that the bill is only a recommendation, not the mandate that’s needed.
The bill is a “fraud,” Senate Minority Leader Chuck Schumer, D-N.Y., told reporters. “It’s voluntary. No company has to do it.”
The industry’s Data Center Coalition says consumers won’t get hurt.
The coalition ”supports the underlying goal of the Ratepayer Protection Act, and we appreciate the bipartisan effort that brought it to the House floor….we welcome the chance to work with the Senate and the Administration to ensure all large-load customers and stakeholders pay their own costs for grid upgrades and related infrastructure,” said Michele Nellenbach, the group’s vice president of federal affairs.
Competition with China is an issue. There’s a risk that if U.S. firms slow development, Chinese technology could surpass them.
Sens. Tom Cotton, R-Ark., and Elizabeth Warren, D-Mass., are pushing the Chip Security Act, which would bar advanced American chips from “falling into the hands of China and other countries of concern.”
Sen. Jim Banks, R-Ind., is among the senators from both parties who are sponsoring a bill that would ban the export of advanced AI chips to China.
Sens. Andy Kim, D-N.J., and Pete Ricketts, R-Neb., are seeking assurances that China can’t get the equipment it needs to make the advanced chips.
Trump said last week that while he’s open to talks about AI, he’s unworried about any threat from China.
“We’re about maybe a year and a half up on China,” Trump told Fox News Sunday. “We’re leading, and we’re building tremendous, trillions of dollars’ worth of places. And why should we give that up?”
What’s ahead?
People want assurance that AI will be safe, and “they want the United States to continue to be a leader on innovation. The two are not mutually exclusive,” Kim told reporters last week.
There is much to learn, said Kennedy.
“It is hard to know precisely what to believe about artificial intelligence. There are some things, it seems to me, that we do know,” he said. “We know, for example, that it is a breathtaking technology.”
Kennedy tried to get the Senate to vote this month on his kill switch plan. Sen. Rand Paul, R-Ky., blocked the bill.
Their dialogue on the Senate floor illustrated why it’s so difficult to act. Kennedy wanted quick action. Paul wanted reflection.
“When approaching something as important as innovation, as AI, I think we should make sure we have as many facts as possible before we dictate rules for the whole economy,” Paul said.
Kennedy was frustrated.
“We are not going to get up off of our ice-cold, lazy butts and address this anytime soon,” he said. “It is just not going to happen, and we all know that.”
