Senator Ed Markey (D-MA) unveiled a multi-bill legislative package on Friday under the banner “AI accountability agenda,” targeting automated hiring systems, datacenter environmental impacts, algorithmic bias, and AI chatbot interactions with minors. The Guardian reported on the agenda in an exclusive, noting Markey has already authored close to a dozen AI-related bills during his tenure.
“Every American is entitled to these safeguards. It shouldn’t be limited just by geographic boundaries of the individual states,” Markey told The Guardian. He stressed that a piecemeal approach to AI regulation “would leave too many people exposed.”
The Bills
Automated hiring ban: Employers would be prohibited from primarily relying on automated systems for hiring, firing, and promotion decisions. The legislation responds to growing use of AI screening tools that filter candidates before a human reviewer sees them.
Datacenter certification: Companies proposing or operating datacenters would need Federal Communications Commission certification affirming the facilities “will not harm the public interest,” even before construction begins. The FCC would evaluate effects on air and water quality, noise levels, energy costs, electricity reliability, local ecosystems, and jobs, consulting with the EPA and local zoning boards, according to a preliminary draft shared with The Guardian.
“We need to make sure these datacenters don’t turn into pollution bombs,” Markey said.
Pre-release bias audits: AI developers would be required to conduct detailed, independent audits of potential bias and discrimination before releasing algorithms that make important decisions.
Additional proposals: Every federal agency using, funding, or overseeing AI would need a civil rights office focused on combating algorithmic bias. Healthcare facilities would be required to create human override options for AI decisions. Workers who disagree with AI recommendations would receive legal protections. Companies would report datacenter energy and environmental impacts in a standardized format.
Legislative Prospects
Markey already has momentum on at least one front. The Senate passed his Children and Teens’ Online Privacy Protection Act in March, banning targeted ads to minors and setting strict limits on collecting personal information from children, The Guardian reports.
The broader agenda faces the same headwind every federal AI bill has since ChatGPT launched in 2022: Congress has not passed comprehensive AI regulation. Markey expressed optimism that bipartisan support will eventually materialize. “Ultimately, there will be national solutions that will be put on the books,” he said.
The Agent Governance Gap
The package does not specifically mention AI agents or autonomous systems, but the automated hiring ban and healthcare override requirements apply directly to agent-driven workflows. Any enterprise deploying agents that make hiring decisions, process insurance claims, or triage patient care would fall under these provisions if they become law. The datacenter certification requirement also adds a new compliance layer for the infrastructure powering agent workloads at scale.