By George Patsarikas
IP Attorney, Burris Law
A recent Southern District of New York decision sent a clear message: materials created using third-party AI tools, even if later shared with counsel, may not qualify for attorney-client privilege or work product protection.
Although this case did not arise in the patent context, the ruling serves as a cautionary reminder that using generative AI to draft or refine invention disclosure materials without clear legal safeguards could create discoverable records, undermine confidentiality, and ultimately expose core intellectual property to future challenges.
In United States v. Heppner, No. 25-cr-00503 (S.D.N.Y. Feb. 10, 2026), the defendant used Anthropic’s Claude AI to generate documents related to a federal securities and wire fraud investigation. He later provided those AI-generated materials to his attorneys and asserted attorney-client privilege and work product protection.
The court held that the AI-generated materials were not protected because they were not communications with counsel and sharing them with attorneys after the fact did not retroactively create privilege. The court also found no reasonable expectation of confidentiality in communications with a publicly available AI platform and concluded that the work product doctrine did not apply because the materials were prepared on the defendant’s own initiative rather than at counsel’s direction.
The practical takeaway is straightforward: sending AI-generated material to your lawyer does not automatically make it privileged.
What This Decision Means for Innovators
Inventors may use generative AI to draft invention summaries, refine technical descriptions, explore alternative embodiments, and even suggest claim language. And these AI tools may be used before consulting patent counsel.
But under reasoning like Heppner, these materials generated outside the context of seeking legal advice may not be privileged.
In the patent context, that risk carries real consequences. For example, the AI-generated materials may be considered a “public disclosure” that inadvertently triggers the one-year grace period for filing an application in the U.S. (and effectively prevents meaningful patent protection abroad). Compromised confidentiality can also affect inventorship determinations, prior art defenses, and discovery exposure in future enforcement actions. Convenience at the front end of innovation can become a vulnerability years later.
For innovation-driven companies, AI governance should be part of patent strategy, not an afterthought. Internal policies, inventor training, and engagement protocols should address when AI tools may be used and under what safeguards. Invention disclosures intended for protection are best developed within a structured legal framework from the outset.
As generative AI becomes embedded in R&D workflows, courts are signaling that privilege doctrines will not automatically expand to protect your inventions. Protecting intellectual property increasingly requires protecting the integrity of how the innovation record is created.
About George Patsarikas: George Patsarikas is an IP attorney at Burris Law focused on patent preparation and prosecution. He has deep experience in electrical engineering and computer-implemented technologies, including software, machine learning and artificial intelligence systems, automotive control systems, semiconductors, and communications technologies. George works with inventors, in-house counsel, universities, and companies ranging from startups to Fortune 500 organizations to protect IP, reduce risk, and identify revenue-generating opportunities, while providing strategic patentability, clearance, validity, and infringement opinions aligned with broader business objectives.