Sen. Rand Paul is warning that a new plan to create artificial intelligence officers in federal agencies will become a pathway for President Biden to insert people who want to censor online speech.
The AI LEAD Act to install chief AI officers throughout the government cleared a Senate hurdle Wednesday over the Kentucky Republican’s complaints that the proposal endangers people’s rights and privacy.
“We’re going to put in place something on AI that we think will protect our privacies, but at the same time I think what we’ll end up finding out is the same people that will be appointed are the same people who think it’s OK to take down constitutionally protected speech,” Mr. Paul said at a Senate Homeland Security and Governmental Affairs Committee meeting.
Mr. Paul sought to amend the bill to reevaluate the AI officer plan in two years, to make the Biden administration use taxpayer cash it has already amassed for the plan, and to train the AI officials annually on the risks of censorship and surveillance.
Sen. Gary Peters, the homeland security committee chairman, helped to tank Mr. Paul’s amendment but expressed a willingness to work on the additional training plan for the AI officers.
The Michigan Democrat said Mr. Paul’s proposed two-year window to sunset the bill was too short, adding that he worked with lawmakers to shrink the sunset provision from 10 years to just more than five years.
The homeland security committee advanced the bill with bipartisan support Wednesday. If the AI LEAD Act becomes law, the incoming chief artificial intelligence officers will likely have responsibility for implementing the National AI Strategy that Mr. Biden’s team is crafting.
The White House Office of Science and Technology Policy has said it’s taking a whole-of-society approach to the AI strategy. Mr. Biden on Friday said he’s preparing to take executive action on tech innovation in the coming weeks.
Also last week, Mr. Biden trumpeted voluntary commitments from seven leading artificial intelligence companies where the tech developers agreed to create new tools safely.
Leaders from Anthropic, Amazon, Google, Inflection, Meta, Microsoft and OpenAI visited the White House to spotlight the companies’ promises that the Biden administration has styled as a bridge to additional regulation.
Congress is likewise working on new rules for artificial intelligence, and the AI LEAD Act is far from the only AI-related legislation under development on Capitol Hill.
The Senate Judiciary Committee has held several hearings examining the danger of AI and is preparing to help write new rules for AI.
The committee gathered testimony Tuesday from Anthropic, whose CEO, Dario Amodei, told lawmakers to worry about the AI-powered proliferation of bioweapons.
“Today, certain steps in bioweapons production involve knowledge that can’t be found on Google or in textbooks and requires a high level of specialized expertise, this being one of the things that currently keeps us safe from attacks,” Mr. Amodei said at the hearing. “We found that today’s AI tools can fill in some of these steps, albeit incompletely and unreliably.”
Mr. Amodei, whose firm delves into AI safety and research, said AI systems appear two to three years away from filling in the gaps in bioweapon knowledge.
Senate Select Committee on Intelligence Chairman Mark Warner wrote to Mr. Biden this week urging him to seek consumer-facing commitments from AI vendors rather than wait for Congress to pass laws.
The Virginia Democrat said Congress has the ultimate responsibility for new AI law but he wants the Biden administration to work on issues involving nonconsensual image generation, facial recognition and the potential spread of cyber problems.
• Ryan Lovelace can be reached at rlovelace@washingtontimes.com.
Please read our comment policy before commenting.