AI conversations without attorney direction lose attorney-client privilege protection
Work product doctrine doesn't cover self-directed AI legal research
Privacy policies matter - AI platforms can share data with government
Arizona practitioners must reconsider AI use in sensitive cases
Attorney supervision is crucial for maintaining legal protections
A federal court decision in February 2026 fundamentally changed how legal professionals think about artificial intelligence and client confidentiality. In United States v. Heppner, Judge Jed Rakoff of the Southern District of New York delivered the first major ruling on whether communications with AI platforms receive the same legal protections as conversations between attorneys and clients. The answer was a resounding no – and the implications reach far beyond New York courtrooms into every law office in Arizona and across the nation.
The Heppner Case: When AI Strategy Backfires
Bradley Heppner found himself in serious legal trouble. Facing federal charges for securities fraud, wire fraud, and conspiracy related to his role as chairman of GWG Holdings, Heppner received a grand jury subpoena that would make anyone nervous. Instead of immediately consulting his attorneys, Heppner turned to Claude, Anthropic's generative AI platform, to help outline potential defense strategies.
Heppner fed the AI information he had learned from his lawyers and asked it to help him prepare reports outlining defense strategies and arguments about facts and law he anticipated the government might charge. He later shared these AI-generated documents with his legal team, believing they would be protected under attorney-client privilege and work product doctrine.
The government had other ideas. Prosecutors moved to compel production of these "AI Documents," arguing they weren't protected by any legal privilege. Heppner's legal team fought back, claiming the documents deserved protection because Heppner had created them in anticipation of legal proceedings and later shared them with counsel.
Judge Rakoff disagreed entirely.
The Court's Reasoning: No Confidentiality, No Protection
Judge Rakoff's February 17, 2026 opinion established two critical precedents that reverberate through legal practice today. First, he ruled that attorney-client privilege didn't apply because Heppner's communications with Claude lacked confidentiality. Second, the work product doctrine offered no protection because Heppner acted without attorney direction.
The Confidentiality Problem
The confidentiality issue proved fatal to Heppner's privilege claims. Judge Rakoff examined Claude's privacy policy, which users must accept to use the platform. The policy explicitly reserves Anthropic's right to disclose user data to "third parties," including "governmental regulatory authorities." It also allows Anthropic to collect user inputs and Claude's responses for training purposes.
Free Case Review
No upfront fees. No legal fees unless we recover money for you.
This policy language, the court reasoned, clearly put users on notice that Anthropic "may" disclose data to third parties. Following another recent decision from the same district, Judge Rakoff noted that AI users "do not have substantial privacy interests" in conversations voluntarily disclosed to an AI platform that retains those conversations in normal business operations.
The court distinguished Heppner's AI documents from confidential notes a client might prepare to share with an attorney. The key difference: "Heppner first shared the equivalent of his notes with a third-party, Claude."
The Work Product Doctrine Fails Too
Work product doctrine protects materials prepared by parties or their representatives in anticipation of litigation. However, this protection typically requires attorney direction or involvement. Since Heppner created his AI documents entirely on his own initiative, without any suggestion from counsel, the doctrine offered no refuge.
Judge Rakoff emphasized that the documents didn't reflect counsel's strategy at the time of creation – they represented Heppner's independent efforts to analyze his legal situation using AI assistance.
Implications for Arizona Legal Practice
While Heppner was decided in New York federal court, its reasoning applies to legal practice nationwide, including Arizona. The decision creates several important considerations for Arizona attorneys and their clients.
Privacy Policy Pitfalls
Most AI platforms operating today include similar data-sharing provisions in their terms of service. Arizona practitioners using AI tools for case strategy, document review, or client communications should carefully examine these policies. Platforms that retain user data or reserve rights to share information with third parties may compromise client confidentiality obligations under Arizona Rules of Professional Conduct.
The Supervision Requirement
The Heppner decision suggests that attorney direction and supervision might preserve some protections for AI-assisted work product. Arizona attorneys considering AI integration should establish clear protocols ensuring lawyer oversight of AI interactions involving client matters or litigation strategy.
This approach aligns with existing Arizona State Bar guidance encouraging lawyers to understand and supervise technology tools used in their practice. The key distinction lies between clients independently using AI versus attorneys directing AI use as part of legal representation.
Practical Applications
Personal injury practitioners in Arizona frequently handle cases involving complex medical records, accident reconstruction, and insurance communications. AI tools increasingly assist with document analysis, medical record summarization, and legal research. The Heppner decision emphasizes the importance of maintaining attorney control over these processes to preserve work product protections.
Similarly, criminal defense attorneys must be particularly cautious about client communications involving AI platforms. Arizona's broad discovery rules in criminal cases could expose AI-generated defense materials to prosecution review if they lack proper privilege protection.
The Future of AI in Legal Practice
The Heppner decision doesn't prohibit AI use in legal practice – it establishes guardrails for maintaining traditional privilege protections. Smart practitioners will adapt by implementing AI tools under proper attorney supervision while avoiding platforms that compromise client confidentiality.
Arizona's legal community should expect additional guidance as courts encounter more AI-related privilege questions. The Arizona Supreme Court's Ethics Advisory Committee may issue formal opinions clarifying ethical obligations when using AI tools in client representation. Additionally, the State Bar of Arizona provides practical guidance and best practices documents on this topic.
Meanwhile, AI developers are likely to create more privacy-focused platforms designed specifically for legal use, potentially including features that better preserve attorney-client privilege and work product protections.
Frequently Asked Questions
Does the Heppner decision ban lawyers from using AI tools?
No, the Heppner decision doesn't prohibit AI use in legal practice. The court ruled that unsupervised client use of AI platforms loses privilege protection, but attorneys can still use AI tools under proper supervision and with appropriate privacy safeguards. The key is maintaining attorney control over the AI interaction and ensuring the platform's privacy policy doesn't compromise client confidentiality.
How does this ruling affect Arizona attorneys specifically?
Arizona attorneys must now carefully evaluate AI platforms' privacy policies and data retention practices before using them for client matters. The decision reinforces existing ethical obligations under Arizona Rules of Professional Conduct regarding client confidentiality and competent representation. Arizona practitioners should implement clear protocols for AI supervision and avoid platforms that share data with third parties.
What should clients know about using AI for legal advice?
Clients should never use AI platforms independently to analyze their legal situations or prepare defense strategies without attorney direction. These communications lack privilege protection and could be discoverable by opposing parties or prosecutors. Instead, clients should consult qualified attorneys who can properly supervise any AI-assisted legal work while maintaining appropriate confidentiality protections.
