The legal profession has operated on the same basic economic premise for more than half a century: lawyers bill by the hour, clients pay by the hour, and the meter runs whether the work takes brilliance or brute force. That model is now facing an existential threat — not from regulatory reform or client revolt, but from artificial intelligence that can perform in minutes what once took associates days.
Anthropic’s general counsel, Jeff Bleich, didn’t mince words when he told Business Insider that the billable hour is “doomed.” Bleich, a former U.S. Ambassador to Australia and longtime partner at Munger, Tolles & Olson, has spent decades inside the machinery of elite legal practice. He knows how law firms make money. And he’s now sitting inside one of the companies building the technology that could fundamentally rewire that equation.
“The billable hour has been dying for a long time,” Bleich said. “AI is just going to accelerate that death.”
His argument is straightforward. When an AI system can draft a contract, summarize thousands of pages of discovery documents, or produce a first-pass legal memorandum in a fraction of the time a junior associate would need, the hourly billing model collapses under its own logic. A client won’t pay $500 an hour for work that a machine completed in 90 seconds. The value of legal work, Bleich contends, will increasingly be measured by outcomes and expertise — not by how long someone sat at a desk.
This isn’t theoretical anymore.
Across Big Law, firms are quietly integrating AI tools into their workflows. Allen & Overy was among the first major firms to deploy an AI assistant built on OpenAI’s technology, a tool called Harvey that helps lawyers with contract analysis, due diligence, and litigation research. Latham & Watkins, Davis Polk, and dozens of other Am Law 100 firms have followed with their own AI pilots and enterprise agreements. The adoption curve has been steep and largely invisible to the public — but inside these firms, the conversations about what AI means for staffing, pricing, and profitability are anything but quiet.
Bleich drew a pointed historical analogy in his remarks to Business Insider, comparing the current moment to IBM’s transformation in the 1990s. IBM, once a hardware company that sold mainframes and PCs, recognized that the real money was shifting to services and consulting. It reinvented itself. Law firms, Bleich suggested, face a similar inflection point. The firms that cling to the billable hour as their primary revenue model risk becoming the legal equivalent of companies still trying to sell mainframes in a cloud computing world.
The comparison is apt in more ways than one. IBM’s transformation was painful. It involved massive layoffs, a wholesale rethinking of corporate identity, and years of uncertainty before the new model took hold. Law firms attempting a similar pivot will face their own version of that pain — particularly when it comes to the armies of junior associates and contract attorneys who have traditionally performed the high-volume, low-complexity work that AI now handles competently.
But here’s the tension. Law firms are partnerships, not corporations. Their governance structures make rapid strategic pivots extraordinarily difficult. Partners who built their books of business on hourly billing have little incentive to dismantle a system that has made them wealthy. The economics of Big Law are built on a leverage model: hire many associates at relatively high but fixed salaries, bill their time at a steep markup, and distribute the surplus to equity partners. AI threatens the foundation of that pyramid.
Some firms are already experimenting with alternative fee arrangements — fixed fees, success fees, subscription models, and hybrid structures that blend hourly work with flat-rate components. These aren’t new concepts. Corporate clients have been pushing for them for years, especially since the 2008 financial crisis forced general counsels to scrutinize outside legal spending more aggressively. What’s new is that AI gives firms the operational capacity to actually deliver on those models profitably. If a firm can use AI to complete a due diligence review in one-tenth the time, it can offer a fixed fee that undercuts competitors while still maintaining healthy margins.
The implications for legal talent are significant. According to a report from Goldman Sachs published in 2023, legal work is among the professions most exposed to automation by generative AI, with an estimated 44% of legal tasks potentially automatable. That doesn’t mean 44% of lawyers will lose their jobs. It means the nature of legal work — and the skills that command premium compensation — will shift dramatically toward judgment, strategy, client counseling, and courtroom advocacy. The rote work disappears. The thinking work remains.
Bleich acknowledged this shift directly. He told Business Insider that the lawyers who thrive will be those who can “add value that a machine can’t” — the ones who understand not just the law but the client’s business, the regulatory environment, and the human dynamics of negotiation and dispute resolution. In other words, the partner who wins a bet-the-company trial will still command enormous fees. The associate who spent 14 hours reviewing boilerplate NDAs will not.
Not everyone in the legal industry shares Bleich’s certainty about the billable hour’s demise. Some managing partners argue that AI will simply make hourly billing more efficient — allowing lawyers to bill fewer hours per task but take on more matters simultaneously, keeping revenue stable. Others point out that clients in high-stakes litigation and complex regulatory matters still want — and will pay for — the assurance that a human lawyer has personally reviewed every document, every argument, every footnote.
There’s merit to that view. For now.
But the trajectory is clear. Thomson Reuters reported in early 2025 that adoption of generative AI tools among law firms has more than doubled year over year, with the largest firms leading the charge. A survey by the firm found that 78% of Am Law 100 firms had either deployed or were actively piloting AI tools for legal research and document drafting. And the tools themselves are improving at a pace that makes last year’s capabilities look primitive.
Anthropic’s own Claude model, which Bleich’s company develops, has shown particular strength in tasks requiring careful reading comprehension and nuanced analysis — precisely the skills that legal work demands. Competitors including OpenAI, Google DeepMind, and a growing roster of legal-specific AI startups like Harvey, Casetext (now owned by Thomson Reuters), and EvenUp are all racing to capture the legal market. The technology is not waiting for the profession to reach consensus on how to respond.
The corporate clients who pay the bills are also accelerating the shift. General counsels at major companies have begun mandating that their outside firms use AI tools, viewing it as a matter of cost efficiency and competitive necessity. Some are going further, building their own in-house AI capabilities to reduce dependence on outside counsel altogether. The message from the buy side of legal services is unambiguous: deliver more value, faster, for less money. Or we’ll find someone who will.
This dynamic creates a classic innovator’s dilemma for Big Law. The most profitable firms — the ones billing $2,000 or more per hour for senior partner time — have the most to lose from a transition away from hourly billing. They also have the most resources to invest in AI. Whether they use those resources to protect the old model or build a new one will determine which firms dominate the next decade of legal practice.
Bleich’s framing of the IBM analogy suggests he believes the answer is already determined. Companies that recognized the shift early and adapted survived. Those that didn’t became footnotes. The legal profession, for all its tradition and institutional inertia, is not immune to the same forces that have reshaped every other knowledge-work industry.
So what does the post-billable-hour law firm actually look like? Probably something closer to a management consulting firm than a traditional legal partnership. Pricing based on the complexity and value of the engagement, not the hours consumed. Smaller teams of highly skilled lawyers supported by AI systems that handle research, drafting, and analysis. A premium on strategic thinking and client relationships rather than document production. And, inevitably, fewer lawyers overall — particularly at the junior level, where the traditional apprenticeship model may no longer make economic sense for firms that can train an AI system instead of a first-year associate.
That last point is perhaps the most uncomfortable one for the profession. Law schools produce roughly 35,000 graduates per year in the United States. Many of them enter Big Law expecting a well-worn career path: grueling associate years, a shot at partnership, and eventually a share of the profits. If AI compresses the bottom of that pyramid, the math changes for everyone — law schools, students carrying six-figure debt, and the firms that have relied on a steady supply of affordable junior labor.
None of this will happen overnight. Legal institutions move slowly, regulated industries move slower, and the courts themselves are only beginning to grapple with questions about AI-generated legal work. But the direction of travel is unmistakable. The billable hour, that peculiar invention that turned legal expertise into a commodity measured in six-minute increments, is losing its grip on the profession that created it.
Jeff Bleich isn’t the first person to predict its demise. He may, however, be among the first to do so from inside a company with the technology to actually make it happen.
WebProNews is an iEntry Publication