AI Law Moratorium Faces Senate Hurdles, and Federal Judge Limits Chatbots Free Speech — AI: The Washington Report

• The proposed 10-year ban on state AI regulations that passed the House is now headed to the Senate, where it faces an uncertain future.
• Several Republican Senators have come out against the proposed moratorium on policy grounds, arguing that it represents a “giveaway to Big Tech” and that states should be allowed to regulate AI until federal AI regulations are passed.
• Some lawmakers are also concerned that the moratorium may not survive the Byrd Rule, a Senate rule that limits reconciliation packages to only budgetary issues. On May 21, a federal district court judge in Florida ruled for the first time that an AI model’s output does not constitute protected speech under the First Amendment, denying an AI company’s motion to dismiss a claim over alleged harms caused by its AI chatbot. The judge’s ruling represents one of the first times a federal judge has ruled on the constitutional rights and limits of AI outputs.  
   

Proposed State AI Law Moratorium Heads to Senate

After passing the House as part of the reconciliation package, as we covered, the proposed 10-year moratorium on state AI laws is now headed to the Senate, where it faces potential challenges. Several Republicans have come out against the proposed moratorium on policy grounds and given the slim Republican majority in the Senate, Republicans can only afford to lose a few votes to pass the measure. Furthermore, the Senate Byrd Rule, which limits reconciliation packages to only budgetary issues, may be a potential obstacle in the way of the measure’s passage.

Some Republican Senators support the provision as it aligns with Congress’s constitutional power to “regulate interstate commerce,” according to Sen. Todd Young (R-IN). Sen. Mike Rounds (R-SD) also said “…the concept and the need for [the moratorium] is important.”

However, other Republicans are skeptical about the moratorium. Sen. Marsha Blackburn (R-TN) argued that state regulation of AI is necessary to “protect consumers until federal legislation is passed,” pointing to a recent law in Tennessee protecting artists from unauthorized AI use. This provision poses the concern that protective measures cannot be halted with the preemption of state AI laws. Other Republican Senators, including Sen. Josh Hawley (R-MO) and Sen. Jerry Moran (R-KS), have expressed that a 10-year pause is a “giveaway to Big Tech” and “damaging to our country's economy and national security.”

Democrats are not in favor of the 10-year moratorium, regardless of whether it is passed as part of reconciliation or as standalone legislation. Some Democratic lawmakers are concerned about the moratorium’s compliance with the Byrd Rule, which prevents any matters deemed as “extraneous” from being included in a reconciliation package that is not directly related to the budget. The moratorium would potentially constitute a policy change, as pointed out by Rep. Alexandria Ocasio-Cortez (D-NY). Sen. Ed Markey (D-Mass.) has also publicly vowed to challenge the moratorium and raise a point of order if “Senate Republicans keep the House language in their reconciliation bill.” Ultimately, the Senate Parliamentarian will take a position on whether the proposed moratorium is consistent with and abides by the Byrd Rule.

Outside of the Senate, the moratorium has received general support from several privacy and advocacy groups and tech companies. One large AI company noted that “it’s impossible to ignore that we officially crossed the threshold of a thousand state AI bills.”

Federal Court Rules AI Output Not Protected by First Amendment

On May 21, a federal district court judge ruled for the first time that an AI model’s output does not constitute protected speech under the First Amendment. United States District Court Judge Anne C. Conway of the Middle District of Florida rejected an AI company’s motion to dismiss, allowing the plaintiff’s complaint about alleged harms caused by the outputs of the company’s AI model to proceed. The judge’s decision is significant, as it marks one of the first times a federal judge has ruled on the constitutional rights and limits of AI outputs.

The case arose from the death of a 14-year-old who died by suicide after engaging with the chatbot platform Character.AI, which allegedly generated abusive and sexually exploitative messages directed at the child. The plaintiff, the boy’smother, brought the suit against the AI chatbot, alleging that Character.AI’s outputs “caused the death of her son,” as stated in the court’s order.

Character Technologies, the developer of Character.AI, argued in its motion to dismiss that the “First Amendment protects the rights of listeners to receive speech regardless of its source,” and pointed to “numerous instances where courts have dismissed similar tort claims against media and technology companies to protect the viewers’ and listeners’ First Amendment rights.” The defendant cited Supreme Court precedents to support its position that “the First Amendment protects speech, not just human speakers,” quoting Justice Scalia’s concurrence in Citizens United v. Federal Election Commission that “the First Amendment is written in terms of ‘speech,’ not speakers.”

In her order, the district court judge ruled that “the Court is not prepared to hold that Character.AI’s output is speech,” noting that the “defendants failed to articulate why words strung together by an LLM are speech.” The case now proceeds to discovery, and defendants can refile their motion on the Amended Complaint within 90 days. All defendants must respond to the amended complaint by June 10, 2025.

Judge Conway supported her decision about the AI model’s free speech protections by citing Supreme Court Justice Barrett’s concurrence in Moody v. NetChoice, in which the Justice suggests that content moderation decisions made by AI may receive less constitutional protection than those made by humans. In her concurrence, the Justice drew a distinction between an algorithm that is directed by humans to moderate “hateful” content, and an AI tool handed the reins to moderate such content. “If the AI relies on large language models to determine what is “hateful” and should be removed,” Justice Barrett wrote, “has a human being with First Amendment rights made an inherently expressive choice … not to propound a particular point of view?” According to Judge Conway, “Character.AI’s output appears more akin to the latter [scenario] at this stage of the litigation.”

The judge’s decision comes as courts nationwide grapple with new legal questions raised by the emergence of AI. Where most federal cases involving AI concern copyright issues for AI-generated work, as we’ve covered extensively, this free speech case stands out because it touches on First Amendment limits on AI output.

We will continue to monitor, analyze, and issue reports on these developments. Please feel free to contact us if you have questions about current practices or how to proceed.

Subscribe To Viewpoints