The “death of the billable hour” has been predicted so many times its starting to feel less like a forecast and more like a ritual in legal tech commentary. Every few years, a new wave of technology arrives and someone declares that the hourly model is finally on its way out.
Artificial intelligence is once again forcing that conversation. But the reality emerging inside law firms looks less like a sudden collapse and more like a slow, structural shift. To understand why, it helps to step back and ask a simple question: what problem was the billable hour actually solving?
For decades, the answer has been uncertainty.
When a firm begins working on a matter, it rarely knows exactly how much work will be required, how complex the issues will become, or how many lawyers will ultimately need to be involved. The billable hour absorbs that uncertainty. If the matter turns out to be straightforward, the client pays less. If it becomes complex, the firm is protected.
Kyle Poe, VP of Legal Innovation at collaborative AI platform Legora, believes that dynamic is beginning to change.
“You have to remember what the billable hour was designed to do,” Poe said in a recent conversation with The Legal Wire. “It deals with uncertainty. At the beginning of a matter, lawyers often don’t know how much work is going to be required. Hourly billing effectively passes that risk onto the client.”
Artificial intelligence, he argues, is gradually reducing that uncertainty in parts of legal work. And when uncertainty declines, the logic of hourly billing begins to weaken.
From efficiency to predictability
Most of the early conversation around AI in law has focused on efficiency: faster research, quicker drafting, fewer hours spent on repetitive tasks. But Poe believes efficiency is only part of the story. The deeper impact of AI lies in something subtler but more consequential.
Predictability.
“As you move from human labour to more tasks being handled by AI,” Poe explained, “you start to get much greater predictability about what’s actually involved in the work.”
That shift has implications that go far beyond time savings. When firms can predict the effort required for certain categories of work, they can begin to price that work differently. Historically, fixed-fee arrangements have been difficult for law firms precisely because of uncertainty. Pricing a matter upfront requires guessing how complex the work will become. If the estimate is wrong, the economics quickly fall apart. In practice, this meant many alternative fee arrangements ended up looking suspiciously like discounted hourly rates rather than genuine pricing innovation.
AI changes that equation where work becomes repeatable.
“As firms gain more predictability about what’s going to be involved,” Poe said, “they’re better equipped to rationally price the work.”
According to Poe, this change is already visible in areas where tasks repeat at scale. He points to early-stage due diligence and first-pass document review as examples. These are processes where AI-assisted workflows can produce consistent outputs, allowing firms to better estimate both time and cost.
“When you have repeatability,” he said, “there’s always the opportunity for technology to streamline the work and create more predictability.”
That predictability, in turn, opens the door to different pricing structures.
The billable hour isn’t disappearing, it’s evolving
Despite the dramatic rhetoric often associated with legal AI, Poe does not believe the billable hour is about to disappear.
“I don’t think we’re going to see the death of the billable hour entirely,” he said. “What we’re seeing instead is the continuation of trends that already existed before generative AI.”
Those trends include a steady rise in alternative fee arrangements and a gradual decline in matters billed exclusively by the hour. AI is likely to accelerate that trajectory rather than replace it outright, and one of the clearest examples of this is the emergence of hybrid pricing models. In these arrangements, predictable tasks handled through technology are priced as a fixed component, while higher-judgment work remains billed hourly.
The logic is straightforward. If a portion of the work is largely deterministic, for example, running a defined workflow or AI process, it can be priced accordingly. But strategic analysis, negotiation, and complex legal judgment still involve uncertainty, making hourly billing more appropriate.
Yet even these hybrid models come with complications.
“The trouble with moving away from the billable hour is trust,” Poe said. “In hourly billing, clients can see what work was done. With fixed or hybrid arrangements, there’s always a question about whether the client is getting good value.”
That is why hybrid pricing tends to emerge first in established client relationships. According to Poe, where trust already exists, firms and clients are more comfortable experimenting with different structures. Its in newer relationships, however, where those conversations are harder.
Clients are pushing, but firms are competing
For years, both clients and law firms talked about alternative fee arrangements without making much progress toward them.
“Everybody liked to talk about alternative fee arrangements,” Poe recalled. “But in practice those discussions often just ended up as discounts on hourly rates.”
AI introduces a clearer commercial trigger. As the capabilities of AI become better understood, some clients are starting to question why certain tasks should still be billed by the hour. In response, Poe expects that they will increasingly be embedding expectations about technology use into outside counsel guidelines.
“As it becomes widely known that AI can handle certain tasks well,” Poe said, “clients are going to be less willing to pay an army of associates to do that work manually.”
That pressure creates a choice for firms. They can continue performing the work as they always have, absorbing the time that clients no longer wish to pay for. Or they can rethink how the work is delivered and priced.
Interestingly, the biggest opportunities for pricing innovation may not lie with existing clients at all. According to Poe, the most natural moment to introduce new pricing structures often comes when firms are competing for new mandates. “When you’re trying to win new work,” he explained, “that’s when you can disrupt competitors by offering a different pricing structure.”
In other words, AI-enabled pricing may emerge first not through negotiation with existing clients, but through competition between firms.
The real barrier isn’t pricing , it’s capability
If the shift toward new pricing models is still relatively slow, Poe believes the main obstacle is not resistance from clients, but capability within law firms themselves.
“The biggest barrier is simply understanding what AI can actually do,” he said. Before firms can rethink pricing, they need to understand where AI genuinely changes the economics of legal work. That requires both broad adoption across the firm and deeper experimentation within specific practice areas.
“There’s a need to upskill lawyers across the firm,” Poe said. “But there’s also a need to go deep in particular areas to build reusable workflows that automate larger pieces of work.”
Until those capabilities exist, conversations about alternative pricing remain theoretical.
Cultural hesitation also plays a role. Many firms are understandably cautious about disrupting business models that have sustained them for decades. But Poe sees that hesitation as temporary. “Ultimately firms won’t really have a choice,” he said. “If they don’t think about how to disrupt their own business, their competitors definitely will.”
Collaboration as infrastructure
Another dimension of the AI conversation that Poe believes deserves more attention is collaboration. Law firms exist, in part, because legal work benefits from shared expertise. Lawyers review each other’s reasoning, cross-check analysis, and collectively refine arguments.
AI does not remove that dynamic. If anything, it reinforces it.
“With AI, trust but verify is absolutely the name of the game,” Poe said. “Lawyers need the ability to check AI outputs and review each other’s work.” Collaborative platforms help enable that process by allowing multiple lawyers to review workflows, inspect outputs, and build shared institutional knowledge.
Collaboration also extends beyond the firm itself. Today, much of the interaction between law firms and their clients still occurs through fragmented channels, including emails, calls, and documents passed back and forth. “There’s a lot of inefficiency in the firm–client relationship,” Poe noted. “Creating a shared interface can make it easier for work to move between firms and clients.”
That transparency may ultimately support pricing innovation as well. When workflows are visible and structured, conversations about cost, scope, and timelines become easier to manage.
AI may reshape competition between firms
AI is also likely to reshape the competitive landscape among law firms themselves. Large firms clearly possess advantages: larger budgets, deeper pools of expertise, and greater access to proprietary data. But smaller firms may benefit too.
“AI can be a force equaliser,” Poe said.
Without the overhead of large associate teams, smaller firms can sometimes adopt new workflows faster. In some cases, AI may allow them to compete for work that previously required much larger teams. “We’re already seeing smaller firms become more competitive,” Poe said, pointing to firms using AI to handle tasks that once demanded extensive associate labour.
The result could be a more dynamic legal market, where technological capability matters as much as traditional scale.
Pricing certainty (not time)
The broader shift underway in legal services may ultimately revolve around a simple change in emphasis. Instead of charging for time spent, firms may increasingly charge for certainty delivered.
AI does not eliminate the need for legal expertise. But it does make it harder to justify billing clients for work that has become predictable and repeatable. Tasks such as complex legal strategy, negotiation, high-risk judgment calls, which remain uncertain, may become even more valuable. But the predictable parts of legal work will increasingly be priced as predictable, which brings the conversation back to the billable hour.
It may not disappear, but it may increasingly be reserved for the kinds of work where uncertainty, and human judgment, genuinely remain.
