Unlock a focused set of AI strategies built to streamline your work and maximize impact. This guide delivers the practical tactics and tools marketers need to start seeing results right away:

Stay ahead of the curve with these top strategies AI helped develop for marketers, built for real-world results.

Download the Free Report

Each edition delivers one clear, evidence-backed idea you can use. This week: Using AI Could Nullify Attorney-Client Privilege

For four hundred years attorney-client privilege behaved like a legal status. It attached to client work automatically and traveled with it. In February 2026 a federal judge in Manhattan turned that assumption inside out, holding that documents a criminal defendant generated with the public version of Claude were not privileged and not work product — even though the defendant had pasted in information his own lawyers had given him (Kumar, 2026). Privilege now behaves less like a status and more like a control that has to be configured, documented, and defended.

The scholarship saw this coming. Tye (2024) laid out the analytic problem in Jurimetrics: Model Rule 1.6 forbids revealing information relating to a representation absent informed consent and typing that information into a consumer generative AI tool is a disclosure to a third party whose terms of service usually allow retention, training use, and human review. Privilege built for a two-party conversation does not survive a three-party one, and the third party is a vendor most firms never vetted.

Seyal (2025) sharpens the point in the University of Dayton Law Review by separating two things that practitioners tend to blur. Privilege is an evidentiary rule that protects specific communications from compelled disclosure in litigation. Confidentiality under Rule 1.6 is the broader ethical duty covering all information relating to a representation, regardless of source or form. A vendor breach usually attacks the second and can incidentally destroy the first — one incident, two failures, one lawyer holding the bag.

DeStefano (2026), in scholarship selected for Northwestern's AI for Law Scholars Conference, argues that closed enterprise AI operating under zero-retention, no-training, no-disclosure terms should not waive privilege, for the same reason that using an e-discovery vendor or a cloud host does not. Her framing is the one worth borrowing: AI is analogous to an associate under the functional-equivalents doctrine or to an accountant under Kovel — a helper working under counsel's direction. Enterprise versus consumer stops being an information technology (IT) procurement preference and becomes the fulcrum of the privilege analysis.

Kumar (2026), critiquing Heppner in the Harvard Law Review, argues that the ruling's categorical exclusion of client-directed AI use is too blunt and that privilege should turn on a fact-dependent inquiry: the AI's role in the relationship, the reasonable expectation of confidentiality, and whether counsel directed the use in anticipation of litigation.

The gap between adoption and governance in solo and small firms is severe. Clio's 2026 survey reported 71 percent of solos and 75 percent of small firms using AI, while 57 percent of solos and 55 percent of small firms had no AI policy at all (Reach, 2026). A separate 2026 legal industry survey found 43 percent of firms with no formal AI policy and no plans to create one, and 9 percent with a written and actively enforced policy (Black, 2026).

The mismatch matters because the tools most common in small firms are consumer products with training-and-retention terms that map cleanly onto Heppner. One associate pasting a client email into a free chatbot to summarize it is a plausible privilege waiver, a Rule 1.6 confidentiality breach, and — after American Bar Association (ABA) Formal Opinion 512 — a supervisory failure under Rules 5.1 and 5.3 (American Bar Association Standing Committee on Ethics and Professional Responsibility, 2024). Small firms carry the full obligation without the compliance headcount that makes it manageable at scale.

Rule 1.6(c) requires reasonable efforts, not perfection, and reasonableness is a process the firm can document. Six moves close the gap this quarter.

The confidentiality perimeter now has two doors — the vendor's terms of service and the help desk phone call — and neither closes with anything you can buy. They close with written procedure and the firm's willingness to enforce it before a court or a state bar asks to see the file.

American Bar Association Standing Committee on Ethics and Professional Responsibility. (2024, July 29). Formal opinion 512: Generative artificial intelligence tools. American Bar Association. https://www.americanbar.org/content/dam/aba/administrative/professional_responsibility/ethics-opinions/aba-formal-opinion-512.pdf

Black, N. (2026). 2026 legal industry report: Trends, benchmarks & insights. 8am. https://www.8am.com/reports/legal-industry-report-2026/

DeStefano, M. (2026). How lawyers can add value in the age of AI. The Practice. Harvard Law School Center on the Legal Profession. https://clp.law.harvard.edu/article/how-lawyers-can-add-value-in-the-age-of-ai/

Kumar, M. (2026, March 23). United States v. Heppner. Harvard Law Review Blog. https://harvardlawreview.org/blog/2026/03/united-states-v-heppner/

Reach, C. S. (2026, May). By the numbers: What surveys show about law firm AI adoption. NC Lawyer. https://www.ncbar.org/nc-lawyer/2026-05/by-the-numbers-what-surveys-show-about-law-firm-ai-adoption/

Seyal, M. (2025). Shielding clients' secrets: The need for a rigorous data security standard in the law practice. University of Dayton Law Review, 50(2), 296–328. https://ecommons.udayton.edu/udlr/vol50/iss2/6/

Tye, J. C. (2024). Exploring the intersections of privacy and generative AI: A dive into attorney-client privilege and ChatGPT. Jurimetrics, 64(3), 309–340. https://www.americanbar.org/content/dam/aba/publications/Jurimetrics/spring-2024/exploring-the-intersections-of-privacy-and-generative-ai-a-dive-into-attorney-client-privilege-and-chatgpt.pdf