The Bottom Line
- A federal judge in Florida allowed most of the claims to go forward in a lawsuit accusing Character Technologies of causing a 14-year-old’s death after he became addicted to an artificial intelligence chatbot, raising questions about the liability of AI developers and the nature of chatbot outputs as “speech.”
- The Court expressed skepticism that the chatbot’s outputs constitute “speech” such that it is protected by the First Amendment, contrasting with a Georgia state court ruling that suggested such outputs could be considered expressive speech.
- Cases such as Garcia v. Character Technologies raise complex questions about speech and the protections afforded to artificial intelligence.
The intersection of technology and law has never been more complex, particularly as artificial intelligence continues to evolve and integrate into daily life. Recent legal battles have highlighted the contentious nature of liability in cases involving AI-driven platforms, raising pressing questions about the responsibilities of developers and operators. An example of this tension is reflected in a recent lawsuit that underscores the potential dangers associated with AI chatbots.
Tragedy Meets Technology
In October 2024, the mother of a 14-year-old boy who tragically died by suicide after months of interacting with an AI chatbot filed a lawsuit in Florida federal court against software company Character Technologies. In Garcia v. Character Technologies,the complaint alleges that the young boy’s interactions with Character Technologies’ chatbot platform, Character.AI, a Large Language Model (LLM), contributed to his mental health decline and subsequent death. The complaint asserts claims of wrongful death, negligence, and deceptive and unfair trade practices against Character Technologies, its founders, and Google, which is accused of contributing to the development and dissemination of the platform.
Defendants’ Motion to Dismiss
Defendants filed a motion to dismiss the lawsuit, arguing, in part, that the First Amendment barred all of Plaintiffs’ claims. They argued, first, that they can assert the First Amendment rights of the chatbot’s users, and second, that the outputs generated by the chatbot constituted “expressive speech”, which users of Character.AI have a right to receive. Essentially, the Defendants asserted the First Amendment rights of the users to receive the chatbot’s “speech.” They compared interactions with the chatbot to an interaction with a video game character, which has previously received First Amendment protection.
The Decision
Notwithstanding the Court’s finding that Defendants can assert the First Amendment rights of its users, the Court rejected Defendants’ attempt to dismiss the case on First Amendment grounds, expressing skepticism that the chatbot’s outputs qualify as “speech” at all. This stance contrasts with an earlier decision this year by a state court judge in Georgia in Walters v. OpenAI, LLC that seemed to suggest that a chatbot’s outputs are entitled to First Amendment protections.
The Defendants’ attempts to compare interactions with Character.AI to interactions with a video game character — which have previously received First Amendment protection — did not sway the Court. The Court noted that its decision “does not turn on whether Character.AI is similar to other mediums that have received First Amendment protections; rather, the decision turns on how Character.AI is similar to other medium.” While humans speak through video game characters to converse with players, the Court was not convinced that chatbots communicate with users in a similar way such that it would constitute speech. The Court must later decide whether the chatbot’s output is expressive such that it is speech.
Why This Matters?
Garcia v. Character Technologies is the latest in a series of legal actions asking courts to define the rights and limitations of artificial intelligence, but it surely won’t be the last. The Garcia court did not determine whether Character.AI’s LLM output constitutes expressive speech. How the Garcia court and other courts address this issue will impact the use of chatbots and other artificial intelligence LLMs.
If courts determine that a LLM’s output is expressive speech protected by the First Amendment, owners and operators of LLMs may be shielded from claims asserting damages caused by that output. On the other hand, if the Court and other courts find that an LLM’s output is not protected speech, there will likely be increased litigation risk for LLM owners and operators, as well as stricter regulations and new requirements for oversight and safety design.
This continues to be an evolving area of law, which is further compounded by the fact that lawmakers and regulators continue to grapple with the implications of AI. This uncertainty requires organizations to stay informed about ongoing legal developments as they navigate this complex terrain. We will continue to closely monitor changes in the law to understand potential risks and ramifications concerning the use of artificial intelligence in business operations.