A&O Shearman Commits US$1.1 Billion to Lawyer-Led AI

In short: A&O Shearman has put a US$1.1 billion price tag on the next five years of its artificial intelligence program. Most of that figure is not a cash cheque for software. It is the value the firm places on requiring every practice to spend at least 3 per cent of its time building specialist AI tools for its own area of law, with AI use now feeding into bonuses and performance reviews. For Australian firms and general counsel, the announcement matters less for the headline number than for what it says about incentives, pricing and who owns the build.

Key takeaways

• On 5 November 2026, A&O Shearman announced a US$1.1 billion, five-year AI commitment, with the majority attributed to lawyers' time rather than external technology spend.
• Every practice must invest at least 3 per cent of its time developing specialist AI tools and services for its area of law.
• AI use now counts toward bonuses and performance assessments for associates, partners and business staff, with reusable team assets weighted differently from personal productivity gains.
• The firm reports fixed-fee income up about 15 per cent over two years, while saying the billable hour will lose importance but will not disappear.
• Peer context includes Kirkland & Ellis's reported US$500 million proprietary AI program over three to four years.

What has A&O Shearman announced?

In its own words, the firm is moving into the next phase of an AI strategy that began with an enterprise generative AI rollout in 2022 and the 2023 launch of ContractMatrix, an AI-powered contract tool developed with Microsoft and Harvey. The five-year program is framed as a firmwide change in how legal work is designed and delivered, and in how people are trained, rather than as another licence purchase.

The US$1.1 billion figure is the firm's own valuation. As Legal Business reported, it is drawn predominantly from lawyers' time commitments. City A.M. coverage puts the total at about £816 million. Without a published cash-versus-hours split, treat the headline as a strategic signal rather than a procurement budget.

What the strategy requires of partners and practices

The operational core is the 3 per cent rule. Each practice must put at least that share of its time into building specialist AI tools and services for its own domain, a deliberate rejection of a single generic chatbot across every desk. After four years of testing, the firm says AI has the most meaningful impact when teams embed deep practice and sector expertise into systems that can then be systematised across an area of law. Practice heads choose whether to dedicate people full-time, spread the work, or mix both, and track hours.

September 2026 maturity metrics from the firm show 96 per cent of partners and 93 per cent of lawyers using its AI tools, with more than 558,000 activities on ContractMatrix and Harvey workflow tools and ContractMatrix product uptake said to have risen ninefold over 12 months. Governance remains lawyer-led: a qualified lawyer takes responsibility for advice regardless of AI's role.

How compensation and incentives change

Time allocation alone rarely changes behaviour. According to reporting that draws on the firm's briefing to the Financial Times, AI usage has, since mid-2026, fed into bonus and performance assessments for associates, partners and business staff. Building precedents, prompts and assets the whole team can reuse is weighted differently from using AI only to speed up individual work.

Global head of AI David Wakeling has described the aim as redesigning the service, not telling people to "use AI more". He has argued it is a mistake to chase the latest model constantly, and that the difference comes from leveraging a team's expertise into specific systems. In the firm announcement, he said: "We want to have more upfront client conversations about how the work, the relationship and the economics can evolve with AI."

Global managing partner Hervé Ekué cast the commitment as a decade-horizon bet: "This significant investment signals where we intend to take the firm over the next decade." He added that the firm that leads will "combine exceptional legal judgment with the speed and scale to apply it consistently, anywhere in the world."

How this compares with peer firms

On headline dollars, Bloomberg Law noted that A&O Shearman's US$1.1 billion tops Kirkland & Ellis's reported US$500 million AI platform spend over three to four years, though the programs differ. Kirkland has emphasised proprietary platform build; A&O Shearman emphasises practice-led tools on its Harvey and ContractMatrix stack, including software-like services sold to clients. Other firms are testing lighter incentive models or foundation-model partnerships. Buy, build and sell can coexist; A&O Shearman is trying to do all three.

What Australian firms and GCs should take from it ?

A&O Shearman has a Sydney office, but the Australian lesson is not "match US$1.1 billion". Mid-market and national firms will not. The transferable ideas are sharper: measure investment in time as well as licences; align bonuses with shared knowledge assets, not only personal productivity; and expect clients to ask how AI changes price. The firm says fixed-fee income has risen about 15 per cent over two years and that the billable hour will lose importance without vanishing. That gives Australian general counsel a concrete Magic Circle reference point for engagement terms.

Buyers should also distinguish generic adoption from specialised delivery. Outside counsel should be asked which workflows have been redesigned, who reviewed the prompts and precedents, and whether any AI-enabled product is licensed separately from advice.

Analysis: the real story is the incentive redesign

Legal AI coverage often treats capital announcements as an arms race scoreboard. This one is more interesting as partnership mechanics. By converting lawyer time into a mandated build quota, and wiring AI into bonus assessments, A&O Shearman is trying to make productisation a partner duty rather than an innovation side project. That is harder to copy than buying another seat licence, and easier to copy than raising half a billion dollars for a proprietary platform.

Open questions remain. The US$1.1 billion valuation rests heavily on an internal price of lawyer time. Tying bonuses to AI use creates measurement risk if activity rises without better client outcomes. And a firm that still earns most of its fees by the hour, while funding AI from billable capacity, lives with a structural tension until fixed fees and product revenue grow. The firms that matter next will show practice-owned tools, incentive alignment and a fee mix that shares AI productivity with clients.

FAQ

How much is A&O Shearman investing in AI?

The firm values its five-year AI program at US$1.1 billion. Most of that figure comes from requiring each practice to invest at least 3 per cent of its time building specialist AI tools, plus technology funding. The firm has not published a detailed cash-versus-hours breakdown.

Does AI use affect A&O Shearman bonuses?

Yes, according to reporting based on the firm's briefing. Since mid-2026, AI usage has fed into bonus and performance assessments for associates, partners and business staff, with reusable team assets weighted differently from personal productivity use.

What should Australian general counsel ask their firms after this announcement?

Ask how AI changes price and scoping, which practice-specific tools sit behind the advice, who reviews AI-assisted work, and whether productivity gains are shared through fixed fees, alternative fee arrangements or product licences rather than only through faster hourly delivery.