State Legislatures Race to Fill the Federal AI Regulatory Void

The National Conference of State Legislatures reports that while state houses are aggressively filing proposals, only a fraction—280 bills—became law by early October. Most of these measures target specific harms, such as deepfakes in political advertising or mandatory disclosures when users interact with AI. While Colorado and Connecticut managed to pass broad regulatory frameworks, most states are opting for narrow interventions in healthcare, education, and digital likeness rights.

This localized approach creates a complex compliance burden for firms operating across state lines. Connecticut Sen. James Maroney argues that waiting for federal intervention is effectively a choice to remain unregulated. He posits that public trust is the primary constraint on AI adoption, suggesting that clear state-level guardrails are essential for long-term viability. Some states, like Utah, are experimenting with regulatory sandboxes to balance oversight with industry flexibility, while others are hiring technical experts to bridge the knowledge gap.

Legislators remain caught between the speed of innovation and the difficulty of drafting durable policy. As NCSL VP Natalie Wood noted, citing Nobel laureate Geoffrey Hinton, the goal of this wave of regulation is not to stop technological progress, but to provide the necessary steering. Despite the high volume of proposals, the persistent divide over whether AI should be governed by Washington or individual states ensures that the current patchwork of rules will define the industry's operational reality for the foreseeable future.