Eyebrow-raising AI amendment passes Senate Commerce committee 

At least one Senate Republican is ramping up a fight with Democrats over the country’s nascent artificial intelligence policy. Sen. Ted Cruz, R-Texas, floated a series of proposed amendments at a Senate Commerce Committee markup this week to scale back the Biden administration’s work to regulate the technology, including last October’s executive order.

While roll call votes on Cruz’s proposals to repeal the AI executive order in its entirety and prevent environmental impact assessments on AI systems failed at the markup Wednesday, another amendment that would dull the teeth of federal government actions to prevent bias and discrimination in AI systems advanced in a substitute via voice vote. 

That amendment, which is now a part of a bipartisan bill that would codify the AI Safety Institute, was quickly met with raised eyebrows from some former Biden White House tech staff on social media. Committee Democrats, however, contend there’s still time to “fix” the bill. 

“Sen. Cruz has made clear — through his public statements and the dozens of divisive, radical amendments filed to bipartisan AI legislation — that he is not for making sure AI is developed in a way to protect consumers from harms, ensure their safety or defend their constitutional protections,” a spokesperson for the committee’s Democratic majority said in a written statement.

The spokesperson continued: “Rather than giving the senator a platform to amplify divisive rhetoric and delay committee progress, the Chair accepted the amendment — knowing there will be many opportunities to fix the legislation, and with the support of the other Republican Senators.” 

As written, the amendment to the Future of Artificial Intelligence Innovation Act of 2024 (S. 4178) would prohibit government actions to promote AI systems being designed equitably, modify input data in systems to prevent disparate impacts on protected classes, conduct assessments for impact on protected classes, and review data inputs to ensure systems aren’t biased or spreading misinformation. The amendment also includes several prohibitions on actions such as promoting one race or sex as superior.

When asked for a comment, a Republican committee spokesperson told FedScoop in an email: “It’s alarming that Democrats now disagree with prohibiting agency rules from suggesting one race is inherently superior to another.” The comment didn’t address the other aspects of the amendment.

Cruz’s amendment and comments during Wednesday’s markup come as the Republican National Committee’s 2024 platform aims to rescind President Joe Biden’s executive order on the technology, saying it “hinders AI innovation.” Cruz pointed to that platform at the hearing and also took aim at companies working with the Biden administration. 

“Big Tech and Big Government get together to protect us rubes and plebes who cannot be trusted. In reality, what is happening here is a tale as old as time: regulatory capture,” Cruz said at the markup.

At least part of the Biden EO does appear to build on the Trump White House’s executive order 13960, which included the creation of federal agencies’ AI use case inventories.

Despite the disagreement, the committee ultimately passed multiple bipartisan AI-focused bills at the markup, including legislation to codify the National AI Research Resource. And Commerce Committee members on both sides of the aisle seemed largely satisfied with the progress of the legislation.

“Yesterday’s mark-up is just the first step in advancing legislation through the Senate. Chair [Maria] Cantwell is committed to protecting people from harm caused by artificial intelligence, including bias and discrimination,” the committee spokesperson said. 

Sen. Todd Young, R-Ind. — who co-sponsors the Future of Artificial Intelligence Innovation Act with Cantwell, a Washington Democrat, and others — praised the passage of the bills in a Friday release, calling them “incredibly important AI policy ideas that will strike the right balance on targeted voluntary standards, international coordination, access to research resources, and public awareness.”

In response to a question about whether he supported the amendment remaining in the bill, a Young spokesperson pointed to the passage of the bill via voice vote and said the senator looks forward to getting the bill through the full Senate. They declined to comment further.

After the hearing, Nik Marda, former chief of staff for the technology division at the Office of Science and Technology Policy during the Biden administration, posted about the amendment on X calling it a “terrible provision.”

Suresh Venkatasubramanian, a former technology adviser to Biden who co-authored the AI Bill of Rights, reposted that, commenting: “This Cruz amendment doesn’t just destroy the bill that Cantwell is trying to pass. It also guts the AI EO, the omb guidance and any future action by the federal govt on AI. And it passed the committee!!!”

Venkatasubramanian was unavailable for an interview.

In an interview with FedScoop, Marda acknowledged that there are still more steps in the process for the text to be “fixed,” but also called the amendment “very dangerous fire to be playing with in the meantime.”

“I really hope that an effort to remove that text later is successful, because as written, this would be extremely damaging — not only to important AI regulatory work — but also eroding 60 years of civil rights law,” Marda said.