Evidence-labeled claim
Human–AI conversations have no categorical legal privilege
This page is a compact epistemic record: what is being claimed, what kind of claim it is, how strong the current evidence is, which research supports or challenges it, and what would justify changing the assessment.
IC-CLAIM-007 legal synthesis supported with qualification research position
Scope and boundary
The claim does not say every AI conversation is discoverable or that no privilege can ever attach. It rejects a categorical privilege assumption.
Why it matters
Users can mistakenly treat an AI assistant as a private cognitive extension even when the law may view the service, records, or provider relationship differently.
Strongest objection
Existing doctrines for agents, experts, work product, private notes, or counsel-directed tools may protect particular uses, and future legislation could create a new privilege.
What would change this assessment
A broadly applicable statute or controlling precedent creating categorical AI-user confidentiality or privilege, or a doctrinal shift treating qualifying AI tools like private notebooks by default.
Supporting research
Source-quality and provenance summary
Supporting dossiers currently connect this claim to 45 distinct cited web sources, including 1 official public-authority and 7 scholarly/preprint sources. Challenging or limiting dossiers connect to 49 distinct sources. Source mix is provenance context, not a vote or truth score.
How source classes are defined · Machine-readable source map
Representative sources cited by supporting dossiers
- www.ilga.gov official public authority
- academiccommons.columbia.edu scholarly or university
- digitalcommons.tourolaw.edu scholarly or university
- digitalcommons.tourolaw.edu scholarly or university
- lawreview.vermontlaw.edu scholarly or university
- library.law.unc.edu scholarly or university
- repository.law.uic.edu scholarly or university
- arxiv.org preprint or working paper
Reviewed document-level source notes
Heppner is a fact-specific federal district-court ruling, not a categorical nationwide rule. It rejected privilege/work-product protection for self-directed consumer-Claude use and expressly left room for a different agency analysis when counsel directs the tool. ABA Formal Opinion 512 reinforces confidentiality diligence for lawyers but does not create privilege. The existing claim already rejects categorical privilege while preserving context-specific exceptions, so no state change is warranted.
United States v. Heppner, No. 25 Cr. 503 (JSR), Memorandum (S.D.N.Y. Feb. 17, 2026)
federal district court memorandum nonbinding trial court decision United States District Court for the Southern District of New York
What it establishes: The court held that this defendant’s self-directed exchanges with consumer Claude were protected by neither attorney-client privilege nor work-product doctrine.
Important limitation: The ruling is fact-specific and nonbinding outside the case. Counsel did not direct the Claude use; the court expressly noted that counsel-directed use might present an agency argument. It does not establish a categorical nationwide rule for all human-AI conversations.
ABA Formal Opinion 512 — Generative Artificial Intelligence Tools
professional ethics opinion nonbinding professional ethics guidance American Bar Association Standing Committee on Ethics and Professional Responsibility
What it establishes: Lawyers using generative AI remain responsible for protecting client information and must evaluate provider security, retention, disclosure, and confidentiality characteristics.
Important limitation: The opinion addresses lawyer ethics under the ABA Model Rules. It does not itself create attorney-client privilege for AI-user communications or determine discovery treatment in every jurisdiction.
See all reviewed source notes →
Challenging or limiting research
Representative sources cited by challenging/limiting dossiers
- media.ca7.uscourts.gov official public authority
- media.ca7.uscourts.gov official public authority
- uscode.house.gov official public authority
- uscode.house.gov official public authority
- www.grassley.senate.gov official public authority
- patents.google.com first party technical or policy
- harvardlawreview.org scholarly or university
- firearmslaw.duke.edu scholarly or university