Artificial intelligence (AI) is quickly becoming the latest excuse for politicians and government officials to crack down on First Amendment rights. Congress, state legislatures, and federal agencies are all increasingly considering – and sometimes passing – measures aimed at limiting or regulating the use of AI in political and policy-focused communications. Yet, these measures frequently reach far beyond actual campaigns for office to place a wide range of speech, from a wide range of speakers, at risk of crushing regulation.

One such proposal is being pushed by the Federal Communications Commission (FCC), where Democratic appointee and Chairwoman Jessica Rosenworcel announced a Notice of Proposed Rulemaking (NPRM) earlier this year that would require groups running political and issue ads to disclose the use of AI-generated content. On September 17, People United for Privacy Foundation (PUFPF) filed comments explaining that the FCC’s proposal is both misguided and unconstitutional.

“For nonpartisan, nonprofit organizations like PUFPF, the proposed disclaimer requirement singles out issue advertising for regulation without any adequate justification. The proposed rule would, in practice, impose a significant burden on the core First Amendment rights of issue advertisers, whose interests the NPRM entirely ignores. Not only does the Federal Communications Commission [] lack the statutory authority to adopt such a rule, but the rule’s content-based trigger is not narrowly tailored to the Commission’s regulatory interest and is therefore unconstitutional,” explain PUFPF Vice President Matt Nese and Counsel Eric Wang.

Regrettably, media coverage of the FCC proposal has largely focused on the impact it would have on federal campaigns. Yet, the language in the proposal itself is much broader: The rules would cover ads that “communicate[] a message relating to any political matter or controversial issue of public importance.” Countless nonprofit organizations, including groups that are legally prohibited from participating in campaigns, routinely speak out about “controversial issues of public importance.” The proposed rules would therefore burden their ability to participate in important debates on topics such as policing, immigration, education, abortion, the environment, health care, and other contested policy and social issues.

In addition to practical concerns with the FCC proposal, the agency also lacks the necessary authority to enact the proposed rules. Nese and Wang elaborate:

“Indeed, the NPRM itself implicitly underscores how the proposed disclaimer requirement is inappropriate for agency action when it cites 39 measures regulating the use of generative AI in political advertising that have already been enacted or are under consideration at the state level. Tellingly, as the NPRM notes, all of those measures have been in the form of ‘legislation.’ The NPRM cites not a single measure that has been enacted by a regulatory agency, even though states generally have campaign finance agencies that regulate state-level political advertising and campaign practices.

The NPRM also betrays its shocking failure to give due deference to the legislative branch by omitting any mention at all of the dozens and dozens of bills that are pending in Congress that would address this very issue. By failing to mention Congress even once in the context of regulating content featuring generative AI, the NPRM reflects an imperial bureaucracy mindset.”

The FCC has faced backlash to its proposed rulemaking from Republicans in Congress as well as commissioners at the Federal Election Commission and two Republican-appointed commissioners at the FCC. Others, however, have encouraged the Commission to move forward, such as House Administration Committee Ranking Member Rep. Joe Morelle (D-NY).

PUFPF will continue to monitor the FCC’s concerning proposal and other efforts to regulate AI that infringe on the free speech rights of nonprofit organizations and their members.