On October 28, senior partners at storied Wall Street firm Weil, Gotshal & Manges announced their recommendation to dissolve the 95-year-old law firm.

Some found irony in the announcement, given the firm’s longstanding leadership in bankruptcy practice (including in tax). Some vented about the timing and substance of the communication—and asked what was next for Weil’s 600 associates, incoming class and summer hires. Some commented on the power of the profession’s elite to swap firms and set their (astronomical) compensation packages.

Few, however, explored what Weil’s collapse might say about the burgeoning role of generative AI in high-end legal practice (one exception). Indeed, some expressly dismissed AI as a cause.

There’s no evidence that AI directly led to Weil’s downfall. But the bigger question is whether AI increases BigLaw’s risk—systemic risk, in the parlance of 2008—of similar “bank runs” involving partners at large law firms. More, below the fold.

It’s an open question whether today’s dynamic moment rivals previous structural shifts in the legal profession, such as those connected to the Great Recession and the dot-com bubble. But if BigLaw has an AI-related structural vulnerability (and it may not), the fault lines might emerge from three mutually reinforcing mechanisms, each of which is brought into focus by Weil’s rapid move to dissolve.

First, increased partner portability. Departures by prominent partners appear to have been the proximate cause of Weil’s wind-down. But if AI reduces these partners’ dependence on organizational infrastructure, then rainmakers may face lower costs of departure when moving their practices. Large law firms traditionally provided hardworking junior lawyers, experienced staff, interconnected IT systems, and the other trappings of elite legal practice. To the extent that AI allows exceptionally productive partners to do more with smaller teams and less infrastructure, the value of these partners’ practices becomes more concentrated in the principal’s client relationships and judgment. And those travel with the human, rather than staying with the firm.

Second, weakened institutional moats. The leading theory of law-firm investment in AI appears to be that proprietary content yields a durable competitive advantage. Firms can leverage decades of deal documents, client histories, and internal materials to augment frontier models—and differentiate those firms’ services from competitors’. Kirkland & Ellis has committed $500 million to develop AI capabilities, including a bespoke AI platform built with Palantir Technologies, while other firms have built proprietary systems using OpenAI and Anthropic’s models.

Weil also was pursuing this approach. In August, the firm announced a “strategic collaboration with Google Cloud to deploy Gemini Enterprise for Legal.” The firm’s in-house AI development group also launched a “proprietary AI-powered litigation intelligence platform.” Just a week before Weil’s wind-down announcement, the firm unveiled a 1L summer program in which law students would help develop AI-enabled legal services. The values connected to these initiatives likely resonate with legal professionals: “thoughtful, decentralized innovation.” The problem may have been that things got too decentralized too quickly.

The peril is that content-based moats may prove less protective than firms anticipate. The value of institutional knowledge may depend largely on the people who know how to use it, as well as the client relationships that the knowledge supports. If firms’ internal work product proves fungible—there’s no secret sauce in that credit agreement markup—then departing rainmakers can operate their practices using any other comparable knowledge base. Generative AI works well with data troves, and it’s possible that sufficient volume (rather than specific quality) may provide all the leverage that high-performers need. If so, the nature of legal content moats requires fundamental rethinking.

Third, correlated instability. The first two mechanisms may affect firms throughout BigLaw, rather than just troubled partnerships. AI may enable increased competition for elite talent, reduce the frictions traditionally associated with practice moves, and—through massive up-front capital outlays—add to firms’ cost pressures. At individual firms, this squeeze may not be evident until, well, it is.

For the legal industry, external shocks may not trigger mass movements of law firm partners, since the attractiveness of partners’ outside opportunities depends on firms’ relative profitability. In an economic downturn, all ships suffer an ebbing tide. The advent of generative AI didn’t weaken firms across the board; indeed, its effect probably operates in the other direction. But AI may reduce switching costs for human capital in ways that open the door to more-frequent bank-run events that in turn could disrupt the orderly flow of legal services into the market. That’s systemic risk.

Things could, of course, break the other way. Proprietary AI capabilities may strengthen firm loyalty and make stable institutions more valuable to superstar partners. Junior lawyers may leverage AI in ways that grow business, or generational shifts may spread the wealth more evenly across BigLaw’s next wave of leaders. Perhaps that credit agreement markup turns out to be golden. Weil’s stunning collapse cannot establish which effect will dominate.

But Weil’s rapid deterioration exposes a potential paradox for the legal industry. BigLaw is investing heavily in technology to improve productivity and strengthen market position. And those investments are very likely to yield productivity gains. But what if that technological arms race also makes lodestar partners more mobile—and more willing to move? Firms’ market position could deteriorate, and the industry could become less stable, even as lawyers’ productivity grows. How firms respond to this instability, especially alongside improvements in lawyerly work product, will shape the legal profession’s landscape going forward.

Related TaxProf Blog coverage:

• Not Too Big to Fail (Oct. 30, 2026)
• Is AI a Substitute for or a Complement to Billable Hours? (Sep. 19, 2026)
• BigLaw and AI Legal Workflows (Apr. 28, 2026)
• Vibe Coding for Lawyers and Law Students (Apr. 25, 2026)
• Staying, Leaving, and Giving Back in the Professions (Jan. 31, 2026)
• Notes on Lawyers, AI, and Money (Nov. 18, 2025)
• Book Review: Rumors of the Death Of BigLaw Are Greatly Exaggerated (Oct. 11, 2023)