Wide-ranging risks created by artificial intelligence (AI) loom over a Congress that appears cautious to regulate as lawmakers head into a monumental election year.
After a year filled with hearings and forums discussing the risks and benefits of AI as the technology expands into critical sectors, experts say it is time for Congress to act — while some warn lawmakers are already behind.
“If there’s no action before the end of this Congress, I think there’s going to be a perception that they failed,” said Dewey Murdick, executive director at Georgetown’s Center for Security and Emerging Technology.
“It’s going be a super tough year to do anything, but I think the pressure is on pretty high to do something,” he added.
Senate Majority Leader Chuck Schumer (D-N.Y.), leader of the chamber’s bipartisan working group on AI, has avoided setting a timeline for expected AI legislation. The leader initially promised new AI rules would take “months,” not “years” in June when the AI working group outlined its approach to crafting regulations.
But when Congress left at the end of last year, lawmakers hadn’t moved on any proposals put forward on AI rules. The group also includes Sens. Martin Heinrich (D-N.M.), Todd Young (R-Ind.) and Mike Rounds (R-S.D.).
Schumer spokeswoman Allison Biasotti said in an email that “conversations continued over break between the AI gang staff and senators directly.”
“The success of the AI Insight Forums led to many recommendations and we aim to produce a framework— working in tandem with the respective senate committees — in the near future,” Biasotti added.
As Congress continues to mull an AI framework, the tech industry is steaming ahead with AI advancements. Analysts at Wedbush said in a report earlier this month that 2024 will be the “year of AI,” as expanding use cases for AI “are exploding globally.”
The potential impacts of a guardrail-free AI future are becoming more apparent as the technology expands, and are highlighted in part by court cases that challenge whether AI companies are stealing work from creators to build their large language models.
Darrell West, senior fellow at the Brookings Institution’s Center for Technology Innovation, said Congress is “way behind on AI regulation.”
“AI is being deployed in virtually every sector. It’s making decisions that affect people. There already are a number of problems that are emerging but there’s been virtually no Congressional legislation,” West said.
Even AI critics who have been encouraged by the Senate’s attention say lawmakers need to step up the pace.
Landon Klein, director of U.S. policy at the Future of Life Institute, which organized an open letter in March calling for a sixth-month pause on certain advanced AI training, said the organization has been “pretty pleased” since the letter at Congress’s actions.
Relative to their “typical pace,” Klein said Congress has moved “surprisingly quickly” on the issue.
“That being said, now that they’ve done a lot of the prep work, including Insight Forums to sort of coalesce around these ideas, it’s time in 2024 to put them into action and get them rolling,” Klein said.
Lawmakers, however, are attempting to balance AI regulation with leaving room for innovation.
Schumer has urged lawmakers to avoid moving “too fast“ on regulations. He warned in September that doing so could “ruin things,” following the Senate’s first AI insight forum, which featured tech CEOs, including Tesla’s Elon Musk, Meta’s Mark Zuckerberg and OpenAI’s Sam Altman.
Industry stakeholders have also warned lawmakers to be wary of overregulating AI in a way that could stifle innovation and limit the U.S. from leading in the race to stay on the cutting edge.
West called that fear a “bogus argument.”
“We are so far from overregulation. I mean, basically, there’s almost no regulation of AI at the very time where it’s being used to make decisions that are affecting people,” West said.
“We’re definitely not acting too quickly. The digital revolution is 20 years old, and we’re still letting private companies make most of the major decisions. So, we just need to start regulating because there’s been very little movement in that area to this point,” he added.
But even experts like Klein who said Congress may not be behind yet, said it is time for lawmakers to act. With the 2024 election on the horizon, though, the timeline is politically more challenging.
“It’s no secret that the election is going to absorb a lot of the attention as the year progresses. So, I think the sooner the better in terms of maintaining focus on the issue and giving it the attention it deserves. Certainly, within the next couple of months it’s essential some language is out and moving through the process,” Klein said.
Alexandra Reeve Givens, president and CEO of the Center for Democracy and Technology, said moving into an election year it will be hard in terms of what Congress will be able to do, “but they set the expectations high and so now is the time to deliver.”
Givens said her biggest concern is that Congress only carries forward with a “light touch” bill that is focused on funding research and innovation, without addressing substantive concerns that “put AI on the national agenda,” such as questions around privacy, security and transparency.
Evan Greer, director of the digital advocacy group Fight for the Future, said Congress is “way behind on regulating AI,” and a major piece of that is the compounding lag on a comprehensive data privacy law in the U.S.
“We’ve been sounding the alarm for years about the dangers of AI-powered surveillance systems like facial recognition, and that’s just the tip of the iceberg. The best thing Congress could do on AI would be to build a time machine, go back about a decade, and pass a strong data privacy law. The second best thing they can do is pass a strong data privacy law as soon as possible,” Greer said in an email.
Unlike the U.S., which has debated and failed to pass a comprehensive data privacy law for years, the European Union’s General Data Protection Regulation went into effect in 2018. Now, the EU is also getting ahead of the U.S. by putting forward the EU AI Act in December
The EU AI Act puts Europe ahead of other lawmakers in setting forward a framework for AI rules. The provisional deal still needs to be ratified by a committee where each country has a representative that is expected to take place early next year before the AI Act would become legislation, Reuters reported.
And, like with privacy, in addition to the EU California and other states are considering AI rules in lieu of federal policy.
“The parallels to what happened and is happening in the privacy space should be lost on none of us,” Givens said.
“There’s a lot that Congress should be learning from that and not replicating mistakes that’s been made in that space,” Givens said.