Blog

Law firms adapt teams for AI but rarely steer on matter profitability

BigHand research 2026: 99% of firms adapt teams for AI, but only 22% factor matter profitability into how they assign work.

· By

A large planning board with a dense grid of pinned task cards on the left and an almost empty zone on the right holding one untouched budget overview; timesheets lie on the table.
Nearly all firms adapt teams for AI, yet only 22% factor matter profitability into how they assign work.Image: IamVera.ai — original editorial illustration

BigHand research of 18 November 2026 shows that almost all law firms are adapting their teams for AI, but only just over a fifth factor profitability per matter into how they assign work. The commercial value of AI is easier to measure when you staff each matter so that freed-up capacity, extra review work, seniority and pricing become visible together.

Software vendor BigHand says it surveyed more than 800 firm leaders and senior professionals in operations, HR and resource management across the United Kingdom and North America. Almost all respondents indicated that AI is changing their hiring priorities and team composition, while only just over a fifth explicitly factor profitability per matter into how they assign work. At the same time, BigHand reports that almost half of firms have already faced the question of why their rates have not fallen despite AI, and that 42% of firms reported requests for alternative fee arrangements linked to AI efficiencies. These figures come from BigHand's own press release and related research.

Our analysis of the findings is that firms are changing their teams because of AI, but rarely steer that change using matter economics. In this way, how a matter is staffed decides whether AI is more than extra capacity. In our assessment this means that the promise of AI does not fail on the technology, but on the way work is distributed.

Why does the staffing of a matter decide whether AI makes money?

Because AI can free up capacity in one place and create additional review work in another. If no one keeps track of where that happens, the benefit disappears into the margin without anyone seeing it. BigHand reports that more than a third of firms assign work mainly on the basis of partner preference, that two thirds have no formal process to prevent underutilisation of lawyers, and that only a quarter routinely track profitability at matter level.

The BigHand talent and resourcing report states that firms should in fact record where AI frees up capacity and where it creates extra control work. That is the difference between a matter staffed on preference and a matter whose staffing is economically justified.

What does the client notice of this on the invoice?

Clients may see the bill without seeing the internal capacity effects of AI. That helps explain the growing pressure to show how AI-related efficiency is reflected in pricing and value. BigHand's pricing analysis reports that 56% of firms face growing client demand for AI-driven efficiencies or transparency around AI use, while only 29% have adjusted pricing structures for AI-augmented work. It also reports that 31% see improved profitability per matter from AI, although pricing and cost-of-resource insight remain underdeveloped.

The commercial treatment of time savings is not automatic: AI may create additional review work, while firms may choose to reflect efficiency through pricing, discounts or retained margin. Anyone who does not keep these apart either promises discounts that do not exist or holds on to value without being able to explain it.

Has the whole market already been changed by AI?

No. In its Law Firm Rates Report 2027, the Thomson Reuters Institute states that AI currently affects firm economics mainly through new legal demand, not through demonstrable internal efficiency. According to that report, firms have not yet materially reduced their use of junior lawyers, most are still using AI experimentally, and almost none have demonstrated AI's return on investment at scale.

That tempers the picture that AI is already wiping out the junior layer. It shifts the question to something more concrete: not whether you use AI, but whether you can show, per matter, what AI did, what review work it called for and what that cost.

What does this mean for executives, partners and firm leadership?

Our analysis: because only just over a fifth of firms factor profitability into how they assign work, making AI value visible requires responsibility beyond the tech department, including the partners who make review and capacity decisions; that is why firm leadership must now make a staffing plan mandatory before engagement, setting out which work AI performs and which control work that calls for. Because 37% of firms primarily distribute work according to partner preference, while only 22% explicitly consider matter profitability when allocating it, those partners should be prepared to explain the absence of measurable AI benefits per matter; that is why they must assign tasks on capacity, cost, development need and matter risk, and record that choice before the work begins. Because almost half of firms are already being asked why rates are not falling, the pricing function must test pricing models that make AI value explainable without assuming that every time saving automatically goes to the client; you have that conversation before the next quote, not after. Because 42% of surveyed firms report receiving requests for alternative fee arrangements linked to AI efficiencies, you must record, per matter, model use and the scope of human review, so that you can substantiate or refute a discount request or an AFA with figures.

As a summary of that analysis, concrete to assign:

  • Make a staffing plan mandatory before engagement that names which work AI does and which review or escalation work that calls for.
  • Assign tasks on capacity, cost, development need and matter risk, not on partner preference alone.
  • Track capacity, profitability per matter and realisation throughout the whole matter, not only afterwards.
  • Record, per matter, which AI was used and how much human review it cost, as you would also do when setting up measurable human oversight of AI advice.
  • Test pricing models that make AI value explainable, and treat an AI performance claim with the same scepticism as when you are only accepting an AI performance claim after your own test.

The broader context of this belongs with the question of what AI in professional practice demands of professionals, and it touches on the lesson that an AI summary does not release a director from their duty of care. The responsibility remains with the person who staffs the matter and defends the invoice.

Sources and references

  1. Law Firms Are Reshaping Teams for AI, but Few Are Managing the Cost of DeliveryBigHand · 2026-11-18
  2. 2026 BigHand Legal Talent and Resourcing ReportBigHand · 2026-11-18
  3. 2026 BigHand Legal Pricing and Budgeting Trends AnalysisBigHand · 2026-11-18
  4. Law Firm Rates Report 2027: The AI impact you may not have been expectingThomson Reuters Institute · 2026-10-28

Sources: The article draws on the BigHand research of 18 November 2026 (press release, talent and resourcing report and pricing analysis) and on the Law Firm Rates Report 2027 from the Thomson Reuters Institute.

← All articles in this topic ← All articles