AI’s Second Wave in Law and Professional Services (Autumn 2025 Edition)
A follow-on paper in the Managing Partner Series
Managing partners are entering a different phase of the AI cycle. Much of the early commentary focused on client pressure, new entrants, platform risk, value capture and talent disruption. Those pressures still matter. What has changed is the maturity of the conversation. The last two months have seen a continued shift in how the legal sector and professional services are adopting, governing and thinking about AI.
1. From experimentation to embedding: what has shifted since late September
The most significant development since the first edition is that Gen AI is no longer confined to pilots and enthusiastic pockets of the firm. Usage is becoming routine across knowledge retrieval, drafting support and matter preparation, even in firms that do not yet have a formal strategy. Awareness is no longer the barrier. Embedding, and value attribution are.
Surveys across law and professional services show that many lawyers now use AI weekly or monthly. The challenge is that the gap between individual use and firm level adoption is widening. Lawyers have moved faster than their firms, often relying on personal workflows or unapproved tools. This has created an uneven landscape where productivity gains exist but remain isolated.
For managing partners, the question is no longer whether AI is being used, but rather whether its use is coherent, governed and connected to firm strategy. The firms that mature in this direction will gain a structural advantage over those that continue to treat AI as background noise.
2. Clients are recalibrating their expectations again
Internal legal teams continue to advance their own use of AI. The shift now is not only that clients are using the tools, but that they are becoming more fluent in how to evaluate them. General Counsel are starting to ask sharper questions about how external counsel verify AI outputs, how quality is maintained and how efficiency gains feed into pricing.
The tone of client engagement has changed. In house teams are increasingly aware that AI changes the economics of legal work. When they ask how firms are using AI, they are now assessing whether a firm can meet the new standard for responsiveness, clarity and value.
What matters for managing partners is that client expectations are no longer passive. The firms that can demonstrate thoughtful integration of AI, underpinned by control, will gain credibility. Those that rely on individual behaviour without firm infrastructure will be exposed in pitches, reviews and panel appointments.
3. Governance has moved from policy to practice
The last two months have made one point unavoidable. Governance cannot be treated as a document. It must become a daily habit.
Regulators, courts and professional bodies have continued to refine their guidance on AI use. The most common themes are verification, confidentiality, supervision and transparency. Yet in many firms these remain theoretical commitments. Informal use of public tools persists. Quality checks vary widely between teams. Partners are unsure how to supervise AI assisted work without doubling the effort they hoped to save.
This is the difference between having rules and enforcing them. Firms now need practical systems that track which tools were used, how outputs were verified and how supervisors signed off the work. Without this, no policy will survive first contact with real work. More importantly, no regulator or client will accept assurances about responsible use without evidence.
Managing partners should treat this as a commercial issue. The ability to show control will soon become a competitive differentiator in the same way cybersecurity and data protection once were.
4. The business case needs a reset. From hours saved to performance
The early narrative around AI was dominated by hours saved. This provided a helpful starting point but has now reached its limit. Hours do not convert into value unless the underlying business model changes. The last two months of research reinforce that most firms still struggle to explain how AI affects their economics.
What now matters is performance. Managing partners need to define the specific operational and financial indicators that AI can influence. These will vary by practice, but common examples include:
- Cycle time to a usable first draft.
- Realisation and write off rates on fixed fee and blended matters.
- Consistency of clause use and error rates at partner review.
- Win rates in competitive tenders.
- Responsiveness and clarity scores in client feedback.
These are the metrics that determine economics and client value. AI will only be meaningful if it is designed to influence them. The most mature firms are already reframing their AI approach around these indicators, rather than seeking proof of efficiency in isolation. They are moving beyond AI is “innovation” to be closely guarded by an “innovation hub”, to AI applies across all elements of the business.
For managing partners, the message is clear. The business case must focus on the firm’s economics and positioning, not on technology for its own sake. Firms that fail to make this shift will report usage but struggle to show benefit.
5. The next talent challenge. Capability, not capacity
Earlier commentary focused on the likely impact of AI on capacity, including shifts in recruitment volumes and the reduction of routine tasks. That remains relevant, but the last two months have highlighted a different challenge, i.e. capability.
Firms now need lawyers and allied professionals who can work effectively with AI. This involves new skills. Reviewing AI outputs, structuring prompts, integrating tools into workflows, and understanding the boundaries between information and advice. These capabilities determine quality and risk. They also shape the client experience, particularly where firms use AI to support communication, summarisation or earlier stage analysis.
The firms that succeed will be those that provide structured training in these capabilities, redesign supervision around AI assisted work, adjust performance expectations and reward, and recognise that true value of non-lawyers. Treating this as a marginal change will put firms behind their competitors. Treating it as a strategic shift in practice will strengthen both quality and culture.
6. Five decisions that will shape 2026
Across all reporting since late September, the following decisions emerge as the most urgent for managing partners. They represent the areas where firms either advance or fall behind.
- Define how AI contributes to your firm’s strategy
Clarify whether AI is intended to improve economics, enhance client experience or expand capability. The firm’s approach will only cohere once this is explicit. - Set firm wide rules for verification and supervision
Move beyond policy and embed practical expectations in matter management. Every partner and supervisor should understand what good verification looks like. - Prioritise no more than five workflows for AI enablement
Avoid scattered pilots. Select workflows that affect economics or differentiation. Build repeatable processes and measure their impact. - Decide how you want to position the firm in front of clients
Clients are becoming sophisticated. Firms need a clear explanation of how they use AI, how they manage risk and how this improves the work delivered. - Invest in capability building
Build AI literacy and oversight skills across the firm. This includes training for juniors and partners, practical exercises, and clear expectations for how AI should be used.
These decisions are not technical. They are strategic and cultural. They determine whether AI becomes embedded or remains fragmented.
Final reflection
The firms that benefit will not be the ones with the most tools. They will be the ones with the clearest strategy, the strongest controls and the most disciplined ability to convert technology into performance.