14
Sources
2,544
Artifacts
6
Frameworks
OPEN
Doctrinal Status

The Central Question

Is AI chatbot output protected speech under the First Amendment, or is it a product subject to standard liability frameworks?

This question remains unresolved following the Garcia v. Character.AI settlement (January 7, 2026). The Conway Order (May 2025) provides the most significant judicial statement: "Court not prepared to hold AI output is speech."

The Central Paradox

IF AI Output = Speech

  • Brandenburg standard applies
  • Requires intent to incite
  • AI cannot form intent
  • → AI harm becomes unregulatable?

IF AI Output ≠ Speech

  • Product liability applies
  • No intent required
  • Strict liability possible
  • → Standard tort framework available

Key Developments

Garcia v. Character.AI - Settled January 7, 2026

The most prominent AI speech case settled before reaching the central First Amendment question. The court indicated it was "not prepared to hold" that AI output constitutes protected speech, but the settlement prevents this from becoming binding precedent.

Impact: Doctrinal question remains OPEN

Raine v. OpenAI - Active (California State Court)

Filed August 2025. Test case examining AI liability for alleged harm. Currently the leading vehicle for establishing precedent on AI speech classification.

See full litigation tracker →

Framework Summary

Framework Classification Strength Key Proponent
First Amendment Speech AI as speaker/publisher MODERATE Volokh, Lemley, Henderson
Products Liability AI as defective product STRONG Nina Brown, Sharkey
Section 230 Exclusion LLMs as content providers STRONG Matt Perault, CDT
Negligence/Duty AI developers have duty of care MODERATE Jane Bambauer
Listener Rights Protection via user rights MODERATE Volokh (J. Free Speech L.)
Hybrid Approach Context-dependent classification EMERGING Harvard Law Review

Detailed framework analysis →

Source Tiers

TIER A = Primary legal sources (cases, statutes, law reviews) | TIER B = Expert analysis (think tanks, academic commentary) | TIER C = News and reporting