By Neven Grigic, Student Attorney
On March 31, the Clinic filed an amicus brief on behalf of three youth organizations in the first case of its kind to consider whether the outputs of generative AI systems are entitled to First Amendment protections. Our brief on behalf of Encode, Design It For Us, and Young People’s Alliance was filed in the Middle District of Florida in the tragic case of Garcia v. Character Technologies, where the mother of a teenage boy who took his own life following a series of emotionally abusive interactions with an AI chatbot has brought suit against its makers for tortious conduct and violations of products liability laws.
Student attorneys Santana Andazola, Jordan Chen, Fynn Fehrenbach, Natalie Phillips, and Neven Grigic drafted the brief under TLPC Director Vivek Krishnamurthy’s supervision. Zoe Leonore Glepa and Telly Scott handled cite-checking and final revisions, while Heidi Mehaffey of Mehaffey P.A. served as our local Florida counsel.
Background
Character Technologies is the developer of an AI chatbot called “Character AI,” which allows users to chat with AI-generated characters—including personas drawn from works of fiction. Plaintiff Megan Garcia seeks to hold Character Technology (and several others) liable for flaws in the design of Character AI, which exposed her 14-year-old son to sexual content that led to addiction, depression, and anxiety. Ms. Garcia alleges that these interactions, including alleged incitements by the chatbot, led her son to end his own life.
Character AI and the other defendants have filed motions to dismiss—a legal tool used to end proceedings before fact-finding when a plaintiff cannot succeed even if all alleged facts are true. The defendants further argue the outputs of its chatbot constitute protected “speech” under the First Amendment, which bars the plaintiff’s claims. Additionally, they contend that users’ First Amendment rights to receive C.AI content would be violated if the plaintiff succeeds.
Our Arguments
On behalf of our clients, TLPC’s amicus brief urged the Court not to dismiss this case of first impression at this early stage of the proceedings, before there has been an opportunity for the development of a factual record. Furthermore, our amicus brief presented several arguments rebutting Character Technologies’ claims that the outputs of its chatbots are protected speech under the First Amendment. Specifically, our amicus brief argues:
- LLM outputs should not qualify as “speech” under the First Amendment because they lack human intent and expressive purpose, instead generating text through statistical methods applied to training data.
- Just as copyright law denies protection to non-human generated content, the First Amendment should not extend to LLM outputs. This extension would prevent courts and legislatures from addressing LLM-specific harms.
- First Amendment issues require factual inquiry that cannot be completed at the motion to dismiss stage. Dismissal without a developed factual record would be premature.
Judge Anne Conway will hear arguments on the defendants’ motion to dismiss Ms. Garcia’s lawsuit on April 28, 2025 at the federal courthouse in Orlando, Florida. TLPC’s amicus brief is available to read and download from the link below.