Senators unveil 'roadmap' for government-funded AI research, regulation

A bipartisan group of senators, led by Majority Leader Chuck Schumer, on Wednesday unveiled a "roadmap" for lawmakers aimed at guiding Congress on regulating artificial intelligence.

They said the recommendations are a critical step as Congress considers legislation to increase innovation and safeguard against negative uses of the rapidly-evolving technology.

The document called for a hefty increase in funding for AI innovation to be worked into Congress' annual funding process -- funding to the tune of $32 billion over several years, an amount recommended by the National Security Commission on Artificial Intelligence.

That money, Schumer said during a news conference will help to "cement America's dominance in AI."

Schumer said that separate funding, perhaps additional billions, will likely also be needed in coming years to address defense and national security concerns related to artificial intelligence.

"Congress faced a momentous choice: We could either watch from the sidelines as AI reshape our world or make a novel earnest bipartisan effort to harness and regulate this industry," Schumer said. "We knew if we did nothing real problems could merge in terms of the lack of maximizing the benefits of AI and the lack of minimizing the detriments of AI."

He said he'll strive to make the effort not only bipartisan but bicameral and that he would soon meet with House Speaker Mike Johnson to discuss AI policy. Schumer said he thinks Johnson is "very interested" in moving forward.

The working group of senators has been focused on artificial intelligence in recent months and it includes Schumer, Sen. Todd Young, R-Ind., Sen. Mike Rounds, R-S.D, and Sen. Martin Heinrich, D-N.M.

Before releasing the "roadmap" Wednesday, the group convened a series of "AI insight forums" over the last several months that featured high-profile names in tech as well as stakeholders across a number of industries. More than 150 experts participated across the nine forums.

The document is the end result of that work. It does not include legislative text, but is meant to help inform lawmakers as they continue to grapple with how best to pass legislation to reign in artificial intelligence. Senate committees have been holding hearings on a number of AI topics relevant to their specific committees and the document encourages the continuation of those discussions.

"We always knew we would have to go to the committees to get the specifics done. There's so many different aspects of AI in so many different areas, that it would take many committees to do it," Schumer said. "We are very now hopeful that the bipartisan momentum that we fostered and the recommendations we made will extend into the committees and their process."

There are a vast number of policy priorities that committees can seize upon, and Schumer said he's hopeful some AI-oriented legislation will be approved by the Senate before the year is out. The Senate, Schumer said, will advance AI legislation as it is ready, and will not wait for a massive package of AI bills to be voted on at one time.

In a major boon to that effort, the Senate Rules Committee on Wednesday advanced several pieces of bipartisan legislation aimed at safeguarding elections from challenges posed by artificial intelligence, including a bill that would ban the use of AI to generate "materially deceptive content" depicting federal candidates.

These so-called "deepfakes" are a focus of the bipartisan report released Wednesday. In addition to backing efforts to regulate fake content concerning elections, the senators also identify a need to regulate destruction of "non-consensual" "intimate" images generated on AI. This issue came into particular focus in January when a sexually explicit AI-generated images of Taylor Swift made the rounds on the internet.

Senators also highlighted in the report the need to consider how to create a federal data privacy framework, how to lead in the adoption of these new technologies in a way that will build up national security, and how ensure that existing laws regulating AI are fully enforced.

They're also said they were excited by the innovative possibilities of AI. Young and Rounds both highlighted the impact that artificial intelligence could have in curing illnesses like cancer and Alzheimer's, and said Congress needs to invest in the innovative technology to avoid medical costs to taxpayers down the line.

"Imagine a world in which we cure cancer, Alzheimer's and other crippling diseases in just a few years' time," Young said. "Imagine a world in which government can be far more efficient, in which we actually figure out how to dramatically cut the healthcare cost curve down."

The senators also say it's important for Congress to consider growing concerns about the impact AI could have on the workforce, including the possibility that workers could be displaced from their jobs by artificial intelligence. They encourage the lawmakers to engage stakeholders across unions and civil society to ensure that workers are trained and retrained to work with AI rather than be displaced by it.

There has not yet been a major piece of legislation regulating AI to pass Congress. While early efforts have seen some evidence of bipartisanship, it's not yet clear what sort of support large scale regulatory efforts would ultimately receive.