Inside the legal industry though, a different kind of tournament is underway — one that won’t be decided by buzzer-beaters, but by preservation letters.
Generative AI tools (e.g., ChatGPT, Copilot, Claude, etc.) have moved from curiosity to core workflow with remarkable speed. A recent Harvard Law School report notes that firms — including AmLaw leaders — are piloting AI tools that compress 16-hour research projects into minutes. [1] That kind of efficiency isn’t optional in today’s market. It’s expected.
But here’s the shift many are only beginning to confront: the legal industry sees generative AI tools as a workflow accelerator. The courts, however, see them as evidence.
That gap — between what saves time and what can be used against you — is where risk lives. And that risk isn’t theoretical. The regulatory and litigation landscape is already shifting in response to how these tools are being used.
Not Your Grandpa’s Discovery Tournament
In college basketball, the modern game looks nothing like it did a generation ago — analytics drive strategy, the transfer portal reshapes rosters, and NIL deals change incentives. Similarly, discovery hasn’t just expanded — generative AI is rapidly changing the rules, the pace, and the players on the court.
In the first few weeks of 2026, lawmakers introduced 78 bills across 27 states focused on chatbots and AI tools.[2] That pace alone should signal something: this technology is not being treated as experimental. It is being treated as consequential.
And litigation trends tell the same story.
In 2025, chatbot wiretap lawsuits under the Electronic Communications Privacy Act were the fastest-growing category of deployer-facing AI litigation.[3] Moreover, in February of 2026 the Southern District of New York (SDNY) ruled that conversations with public generative AI tools — like ChatGPT — can be discoverable, even when those chats relate to legal advice from an attorney.[4]
That decision changed the rules of the game. Discovery is no longer confined to emails, texts, and document drafts. It now may include:
- Prompts entered into AI tools;
- AI-generated outputs;
- Internal summaries created through AI; and
- System logs reflecting usage [5]
The question is no longer whether AI-generated material can enter the discovery arena. The question is whether you are prepared when it does. Below are a few ways to prepare your game plan before the buzzer sounds.
Practical Plays to Keep Your AI Use Defensible
🏀 Draft a Company-Wide AI Playbook
Every championship team runs structured plays. Create a written AI policy grounded in ethics, privilege protection, and discovery awareness.[6]
🏀 Upgrade to Enterprise-Grade AI Tools
Enterprise AI isn’t a nice-to-have — it’s a must for your frontline defense. Pick tools that guarantee confidentiality and don’t store or learn from your data. A good rule of thumb: if it’s free, assume your data protections are limited.
🏀 Coach Your Team
Talent alone doesn’t win tournaments — discipline does. Train your teams on which enterprise grade tools are approved and on what is potentially discoverable when using those tools.
🏀 Coach Your Clients
Your clients are in the tournament, too. Make it clear to them that using public AI tools to summarize a legal email from an attorney or to brainstorm legal matters—can waive privilege, confidentiality, and trade secret protections. [7]
🏀 Integrate AI Into eDiscovery Strategy
Preservation and defensibility belong in the opening tip-off, not as a halftime fix. Add experienced discovery and data forensics teammates to your roster early to make sure data collection is complete and accurate.
Final Four Takeaway
Generative AI is not the problem. Unstructured generative AI use is.
In March Madness, talent gets you to the tournament. Discipline gets you to the Final Four.
In discovery, innovation gets you efficient. Governance gets you defensibility. The teams that make it to the Final Four of this next legal era won’t be the ones that bench AI. They’ll be the ones that run it like a disciplined offense.
[1] https://clp.law.harvard.edu/knowledge-hub/insights/the-impact-of-artificial-intelligence-on-law-law-firms-business-models/
[2] https://www.transparencycoalition.ai/news/chatbot-bill-surge-nationwide-concern-spurs-78-proposals-in-27-states
[3] https://ourtake.bakerbotts.com/post/102mipe/ai-chatbot-regulation-78-state-bills-58-lawsuits
[4] https://www.courtlistener.com/docket/71872024/27/united-states-v-heppner/
[5] https://www.klgates.com/Litigation-Minute-Is-AI-Generated-Content-Discoverable-What-Companies-Need-to-Know-in-2026-2-12-2026
[6] https://www.isba.org/sections/ai/newsletter/2025/03/aiinthesmalllawfirmtheneedforaformalwrittenpolicy
[7] https://www.huschblackwell.com/newsandinsights/heppner-v-claude-the-first-privilege-waiver-by-ai-rulingwhat-lawyers-and-clients-must-know