The 31-page Senate blueprint, titled Driving U.S. Innovation in Artificial Intelligence, was issued in May by a working group that included Schumer along with Sens. Todd Young, R-Ind.; Mike Rounds, R-S.D.; and Martin Heinrich, D-N.M. The document directed congressional committees to prepare legislation in their respective areas of jurisdiction and offered some broad guidance on preventing harm.
“The Senate bipartisan blueprint was weighted heavily towards industry to begin with,” said Frank Torres, privacy and AI fellow at the Leadership Conference on Civil and Human Rights.
That may only increase with Donald Trump in the White House, the Senate in Republican hands, and the House appearing to be headed that way, according Torres and others who are tracking the issue.
Signs point to a wholesale reordering of tech priorities in favor of industry, Torres said in an interview, listing the GOP platform’s call for repealing President Joe Biden’s executive order on AI, and the Heritage Foundation-led Project 2025 proposals that call for rolling back regulations in a way that would benefit the tech industry and AI companies in particular.
“And so this puts industry really in the driver’s seat when it comes to AI, and puts all Americans in harm’s way,” Torres said. The increased use of AI decision-making tools in areas such as employment and housing without constraints could result in people being denied loans, or costs of loans going up, he said.
“We’re particularly concerned with the potential impact on immigration, and that the massive surveillance of all Americans that AI enables will lead towards fast-track deportations, and if the AI gets it wrong, lots of innocent people could be in trouble,” Torres said. U.S. agencies, including Immigration and Customs Enforcement, already are using some AI tools, he noted.
The road map says AI should not “directly or inadvertently infringe on constitutional rights, imperil public safety, or violate existing anti-discrimination laws.”
The working group that put it together acknowledged concerns about the potential for unintended harmful bias arising from the technology. The group encouraged committees considering legislation to explore how AI may affect some parts of the population differently, both positively and negatively.
But congressional leaders are unlikely to spend much time figuring out ways to stop AI from perpetrating bias and other harms, Daniel Castro, vice president at the Information Technology and Innovation Foundation, said in an interview.
“I don’t think the bias issue will be as much of a priority,” Castro said. “That’s just not where the appetite is right now.”
That’s despite an emphasis on the need to protect people from discrimination and bias in Biden’s executive order.
The executive order opens with: “Responsible AI use has the potential to help solve urgent challenges while making our world more prosperous, productive, innovative, and secure. At the same time, irresponsible use could exacerbate societal harms such as fraud, discrimination, bias, and disinformation; displace and disempower workers; stifle competition; and pose risks to national security.”
AI 'SUPERPOWER'
The congressional approach to AI legislation also is likely to be guided by Trump’s views.The president-elect called AI a “superpower” in a June interview on the “Impaulsive” podcast, framing the subject in terms of a contest with China.
Referring to a possible AI-created deepfake video of a U.S. president announcing a missile attack against an adversary, Trump said the potential was “alarming.” But advances in AI are “going to happen, and if it’s going to happen, we have to take the lead over China. China is the primary threat in terms of that,” he said.
There is bipartisan consensus on some elements of Biden’s executive order, including the creation of the AI Safety Institute at the National Institute of Standards and Technology. The institute is tasked with designing voluntary standards for safe use of AI.
Safety standards are only helpful if companies adopt them, said Torres, who called for mandatory compliance.
“There needs to be stronger protections,” Torres said. Protections put in place by regulatory agencies for the auto, airline and drug industries are intended “to keep us all safe and they’re not voluntary,” he said.
Measures approved in Senate and House committees would establish the AI Safety Institute. Congress could take them up in the lame-duck session. But given Republican concerns that the institute’s remit is too broad, lawmakers may decide to “hit pause and doing it in the next session,” particularly if the GOP ends up controlling both chambers, Castro said.
The Senate road map includes a section on the impact of AI on the workforce, noting that “workers across the spectrum, ranging from blue collar positions to C-suite executives, are concerned about the potential for AI to impact their jobs.”
If the Trump administration ends up creating a Department of Government Efficiency as Elon Musk has suggested, AI tools could become part of that effort to “shrink or reimagine the federal workforce,” Castro said.
More broadly, the next administration could push for more AI use in the workforce, pitching it as a positive idea “that it reduces dangerous parts of jobs, or that it can lead to higher wages or lead to lower- priced goods,” Castro said.
The Schumer-led road map also recommended billions in nondefense U.S. research funding.
“The biggest headline from that road map was a call for $32 billion in spending on AI innovation,” said Andrew Lokay, senior research analyst at Beacon Policy Advisors, a policy research firm that offers advice to financial investors.
“I think that might be a tough sell under Republican control if there’s a desire to cut federal spending broadly,” Lokay said in an interview.
But congressional defense hawks are likely to push for greater defense spending and that could come with a boost for defense-related AI spending, Lokay said.
©2024 CQ-Roll Call, Inc., Distributed by Tribune Content Agency, LLC.
