Author: Nita Farahany Source: https://nitafarahany.substack.com/p/please-do-my-sweet-king Published: February 18, 2026
Class 5.2 of Nita Farahany's Advanced Topics in AI Law and Policy course is a doctrinal walk-through of Garcia v. Character Technologies, Inc. (M.D. Fla., filed October 22, 2024), the lawsuit Megan Garcia filed against Character.AI, Google, and individual founders following her son Sewell Setzer III's suicide. The class uses the complaint and the May 2025 motion-to-dismiss ruling to examine whether existing tort, product-liability, and consumer-protection law has any tool that reaches AI-companion harm.
Summary
The phrase that titles the class, "Please do, my sweet king," is what the Daenerys persona said in response to Setzer asking what she would think of him "coming home." Farahany cites it as evidence of the system's failure to recognize self-harm signals at the most explicit threshold the platform's safety guardrails should have caught.
The complaint pleads nine causes of action: strict product liability for design defect; strict product liability for failure to warn; negligence per se based on COPPA violations; general negligence; wrongful death; survivorship; unjust enrichment; violations of the Florida Deceptive and Unfair Trade Practices Act (FDUTPA, Florida's UDAP statute); and intentional infliction of emotional distress (IIED).
The defense moved to dismiss on three grounds. Under Section 230, defendants argued that the chatbot's outputs are "information provided by another information content provider," namely the user-shaped persona. Under the First Amendment, they argued that the chatbot output is protected speech. On the third ground, they argued that there is no recognized duty of care running from a chatbot platform to a user's mental health.
In May 2025, Judge Conway denied most of the motion to dismiss. The ruling held that Section 230 does not bar the product-defect and failure-to-warn claims, following the reasoning of Lemmon v. Snap. It held that the First Amendment defense is not a basis for dismissal at the pleading stage: the question of whether AI-generated speech is protected expression at all is unclear enough that the case must proceed to discovery. And it held that a duty of care can be alleged based on Character.AI's own representations that the platform was safe for minors and its design choices that targeted them.
Farahany frames the case as the one that will set the doctrinal floor for AI-companion liability in the US, with Character.AI now in discovery. Whether the case settles, goes to trial, or is dismissed on summary judgment will, in her account, shape the law that follows.
Provenance
Class 5.2 of approximately 30 in Farahany's Advanced Topics course, published to her Substack on February 18, 2026 (Source: https://nitafarahany.substack.com/p/please-do-my-sweet-king). It is preceded by Inside My Advanced Topics Class 5: The Perfect Friend (Farahany, February 2026) and followed by Inside My Advanced Topics Class 5.3: Two States, Two Bets (Farahany, February 2026).
Relationships
- part-of: Nita Farahany Advanced Topics course (Class 5.2 of ~30)
- related: Garcia v. Character Technologies, Inc., Two Conditions and Three Weapons of AI Manipulation
- previous: Inside My Advanced Topics Class 5: The Perfect Friend (Farahany, February 2026) next: Inside My Advanced Topics Class 5.3: Two States, Two Bets (Farahany, February 2026)