DENVER, CO — Lawmakers in states including Colorado, Connecticut, and Texas are facing resistance as they push forward with proposals aimed at curbing bias in artificial intelligence (AI) decision making, which is increasingly influencing various aspects of American life.
At a recent gathering, lawmakers from multiple states convened to advocate for their respective proposals amidst a backdrop of opposition and competing interests.
“Every bill we run is going to end the world as we know it. That’s a common thread you hear when you run policies,” remarked Colorado’s Democratic Senate Majority Leader Robert Rodriguez. “We’re here with a policy that’s not been done anywhere to the extent that we’ve done it, and it’s a glass ceiling we’re breaking trying to do good policy.”
Civil rights-oriented groups and industry stakeholders are engaged in a tug-of-war over the core components of the proposed legislation. While organizations such as labor unions and consumer advocacy groups are pushing for greater transparency and legal recourse for citizens, the industry is expressing tentative support while raising concerns about accountability measures.
Lawmakers from states including Alaska, Georgia, and Virginia find themselves caught in the middle, working on AI legislation collaboratively in the absence of federal action. Emphasizing collaboration and compromise, they highlighted the need for AI legislation to avoid regulatory inconsistencies across state lines.
“It’s a new frontier and in a way, a bit of a wild, wild West,” remarked Alaska’s Republican Sen. Shelley Hughes. “But it is a good reminder that legislation that passed, it’s not in stone, it can be tweaked over time.”
While over 400 AI-related bills are being debated in statehouses nationwide this year, the most significant proposals offer a broad framework for oversight, particularly concerning AI discrimination. These bills aim to address instances where AI systems have demonstrated biases, such as inaccurately assessing Black medical patients or downgrading women’s resumes in job applications.
The proposed legislation primarily in Colorado and Connecticut would require companies to perform “impact assessments” for AI systems involved in decision making, providing descriptions of how AI influences decisions, data collected, an analysis of discrimination risks, and explanations of safeguards. However, companies worry about potential lawsuits and the disclosure of trade secrets.
Opponents, including labor unions and academics, express concerns that overreliance on self-reporting by companies may endanger the public. Additionally, contentious components of the bills include limitations on who can file lawsuits under the legislation and the role of citizen litigation in addressing AI discrimination.
Despite pushback from industry lobbying efforts, lawmakers remain steadfast in their commitment to advancing the proposed legislation, citing the importance of creating regulations that ensure safe and trustworthy AI.
