This Act authorizes federal agencies to issue binding rules to mitigate violations of existing federal law caused by the use of algorithmic decision-making systems.
Sara Jacobs
Representative
CA-51
The Sectoral AI Governance Act of 2026 grants federal agencies the authority to issue rules regulating algorithmic decision-making systems that are likely to materially contribute to violations of the federal laws they oversee. This framework mandates public input, interagency coordination, and periodic review to ensure rules are consistent, technically sound, and mitigate harm. Agencies must also report biennially to Congress on their rulemaking activities and enforcement actions related to these systems.
New AI Governance Act Gives Federal Agencies Power to Regulate Algorithms: Rules to Roll Out Starting in 2026
The Sectoral AI Governance Act of 2026 is essentially a toolkit for federal agencies to police how companies use artificial intelligence. Right now, there’s a lot of gray area regarding whether old laws apply to new tech. This bill changes that by giving agency heads—like those at the Department of Labor or the FTC—the explicit authority to write new rules if they can prove an AI system is "materially contributing" to breaking existing federal laws. If an algorithm is helping a company illegally discriminate against job applicants or tricking seniors out of their savings, the agency can now step in with specific AI-focused regulations and back them up with civil penalties and fines (Section 3).
The "Materially Contribute" Standard
One of the biggest shifts is the introduction of the "materially contribute" standard. In plain English, this means if an AI plays a significant role in a legal violation—not just a tiny, incidental part—the agency can drop the hammer. For example, if you’re a contractor using a software tool to manage public benefits and that tool starts glitching and denying people healthcare, the agency has to look at how that tech is affecting the "delivery, accessibility, and integrity" of those services. They are required to minimize any "unnecessary adverse effects" on the people relying on those benefits, ensuring that a tech upgrade doesn't leave vulnerable families in the lurch.
Checks, Balances, and the Fine Print
To keep agencies from going rogue or creating a confusing mess of conflicting rules, the bill mandates a heavy dose of teamwork. Before any rule is finalized, agencies have to talk to the Office of Science and Technology Policy and NIST to make sure they actually understand the tech they are trying to regulate. There’s also a mandatory 60-day public comment period where tech experts, small business owners, and regular citizens can weigh in—though there is a bit of a loophole: agency heads can skip this if they decide they already have enough info from previous studies. Every three years, agencies have to look back at their rules and decide if they’re still relevant or if the tech has evolved so much that the laws are now obsolete.
Real-World Reporting and Accountability
Starting two years after the bill kicks in, agencies have to start showing their work through biennial reports. These aren't just dry spreadsheets; they have to include details on the challenges they faced, the types of violations they found, and even a summary of whether the AI systems showed "disparate impacts" or discriminatory effects. This is a big deal for anyone who’s ever felt like a "computer says no" decision was unfair but had no way to challenge it. While the bill gives the government more teeth to regulate, it also leaves the door open for states to pass even stricter rules, as long as they don't directly clash with these new federal standards.