A judge let most claims against a chatbot maker proceed

The design

On 20 May 2025, Judge Anne Conway of the Middle District of Florida ruled on the motions to dismiss filed by Character Technologies, Google, Alphabet, and the individual co-founder defendants in Garcia v. Character Technologies. The order, following oral argument held in April, dismissed the plaintiff's claims against Alphabet without prejudice at her own request, dismissed the intentional-infliction-of-emotional-distress claim without leave to amend, and denied dismissal of the plaintiff's remaining claims, meaning the product liability, wrongful death, negligence and unjust enrichment theories proceed. The court ordered defendants to answer the amended complaint by 10 June 2025.

What the evidence says

The order is a motion-to-dismiss ruling, meaning the court took the complaint's factual allegations as true only for purposes of deciding whether a legal claim was stated, not as a finding that they occurred. On the defendants' central First Amendment argument, that Character.AI's chatbot output is speech its users have a right to receive, the court held it could consider that argument, but wrote it was 'not prepared to hold that Character A.I.'s output is speech,' citing a concurring opinion's distinction between a platform's own expressive choices and output generated by a large language model. The court likewise declined to resolve, at this stage, whether Character.AI is a 'product' for strict-liability purposes.

What it asks of people

The ruling asks nothing of chatbot users directly; it is a procedural stage in litigation between named parties. It asks the surviving defendants to answer the complaint and defend the remaining claims through discovery and, potentially, trial. For a reader, the ruling's value is in showing what a company's stock legal defenses can and cannot yet accomplish: a broad free-speech argument did not end the case at this stage, though it may resurface later.

Privacy and safeguards

The order does not adjudicate any safety feature or privacy practice; it decides only which legal theories may proceed. Its significance for safeguards is indirect: by letting product-liability and negligence claims continue, the court preserved a legal avenue for testing, in later stages of the same case, what a chatbot's design and safety features actually did.

• Does coverage of this ruling treat it as a final verdict, rather than a procedural stage that could still end the other way at trial?
• What would it mean for chatbot regulation broadly if a court eventually holds that Character.AI's output is not speech?
• How does 'taking allegations as true' for a motion to dismiss differ from a jury or judge finding facts after evidence is presented?

The order kept most of the case alive without deciding it. Reporting that reduces the ruling to a single winner or loser skips the qualifier the order itself insists on: nothing here is decided on the merits.