Ten things law firms should be doing about AI now

Alexandru Birsan

Filip & Company, Bucharest

Alexandru.Birsan@filipandcompany.com

Martijn Lesterhuis

Saga, Amsterdam

martijn.lesterhuis@sagalegal.io

In September 2024 the AI & Tech Subcommittee of the IBA Law Firm Management Committee published its first ‘10 things law firms should be doing about AI now’.

With the development of AI within the legal industry moving quickly, we have updated the memo. The objective remains the same: provide an overview of developments and practical steps for law firm leaders with a view to AI.

Although AI has moved past the initial hype stage, there are still many differences in the pace of introduction of AI in law firms, also at country-by-country level. That being said, compared to 2024, AI is no longer being implemented at large national and international firms only, but, in many countries, also within midsize and smaller law firms and, in some countries, even at sole practitioner level.

In addition, we see that law firms’ clients, especially legal departments, have started to implement AI into their legal processes. This will have consequences for the volume and type of work they will request from their law firms. Finally, we also see the first (semi-)public legal organisations, such as courts and complaint organisations, piloting AI.

Below are ten priorities for law firm leaders. The list is not drastically new compared to last year, but rather builds on the previous year’s list. It stresses that firms should not only make a next step, but that they also need to start making a plan for the future with respect to AI.

1. Show leadership at the top and make AI a board responsibility

Showing leadership at the top is becoming even more important than last year, as AI starts to impact law firms at different levels: its IT landscape, policies and governance, investment requirements, client relations, organisation and business model, as well as business operations including marketing, business development, finance, HR and client services.

AI adoption cannot be delegated to IT or innovation managers. Firm leadership must own it, visibly and decisively. Boards should set the vision, allocate budgets and make AI part of firm strategy alongside finance, HR and risk. This includes considering AI applications not only in legal practice, but across business development, marketing, finance, HR and client service to drive holistic firm-wide transformation.

A core team – combining partners, associates, business functions and tech specialists, including dedicated legal AI professionals (eg, legal technologists, legal engineers, AI project managers) – should drive AI adoption, experimentation and integration across the firm. Diversity in seniority and roles is key to ensure solutions serve the entire business.

Firms should make AI leadership a board responsibility and needs a fully empowered partner-level AI leader, supported by a cross-functional core team that includes or has access to specialised legal AI professionals (eg, legal technologists, legal engineers, AI project managers), accountable for implementation and impact. Firms should invest in hiring or upskilling these specialists to build internal AI capability.

Once the strategic decisions are taken, autonomy and support are vital to encourage continued rapid implementation. Critically, AI must be integrated into the firm's overarching business and operational strategies as a horizontal enabler across people, processes, technology and client service. It should not be treated as a standalone innovation track or siloed technology project.

2. Break out of pilot paralysis and create a learning path

More than in other industries, law firms are flooded with demos and pilots for AI solutions. This results in many firms running scattered pilots: testing a tool here, experimenting with prompts there, while solutions are being upgraded continuously. Many firms therefore find it difficult to choose and take the next step. This ‘pilot paralysis’ is holding the actual AI development of the firm back.

Instead of focusing on the latest status of a solution or deciding what the solution should be for the next five years, firms should take a step back and take their own learning path with AI as a starting point. They should question how to realise widespread adoption and use of AI and then decide who will support the journey for the coming year: fitting AI-solutions and external providers, combined with internal champions. In this fast-developing market, it is more important to learn systematically than to endlessly experiment.

Continued testing of new products is healthy, but not at the expense of implementing AI solutions. When law firms try to select an AI solution, we think that organising simultaneous pilots with the same focus group of AI users is crucial for effective learning and comparison. It is also important to keep an open mind to the fact that the final solution may be a portfolio of AI solutions. At this stage of development, the only long-term bad choice is paralysis.

3. The move from individual tasks to workflows with agentic AI

For now, AI is being used for individual tasks (review one or more documents, do a search in a legal source, draft an agreement, review a contract, etc). The next generation of AI is agentic: able to chain tasks, connect with other systems (via Application Programming Interfaces or Model Context Protocol connections) and semi-autonomously manage entire processes.

For law firms, this means moving beyond isolated drafting or research tasks to rethinking end-to-end workflows: due diligence, compliance checks, litigation discovery and contract lifecycle management. The strategic question is not ‘Which tasks can we accelerate?’ but ‘Which workflows can we redesign – and how do we maintain human judgment and accountability?’

Consequently, legal process management will become more important, as well as oversight mechanisms to ensure humans remain in control of judgment and client outcomes. With the arrival of agentic workflows, law firms should think about how they will organise the design of legal workflows in their practice and the impact it will have on their business model. In addition, they should ensure that their lawyers remain in control of both process as well as outcomes, to avoid that the solution make autonomous decisions and mistakes.

4. Institutionalise AI literacy and critical use; focus on AI-enablement

Many firms have already run AI literacy and prompting sessions. While AI solutions are developing fast, law firms should start to embed AI competence into the DNA of the firm.

Looking forward, AI will quickly become more important in legal work processes. Lawyers should develop fluency in working with AI tools: not just understand the best uses and how to prompt, but how to critically assess, apply counterpoints (counter-anchoring) and refine AI output. This is crucial to avoid over-reliance on model defaults that can subtly shape reasoning and weaken judgment.

Law firms should also consider how to add client value beyond out-of-the-box AI solutions. This could include structured ‘productive friction’ – for example, requiring juniors to draft first, compare alternatives and annotate AI suggestions. This helps ensure that AI sharpens, rather than flattens, professional growth. Make AI skills part of continuing education, promotion criteria and daily workflows. Build a culture where AI is a sparring partner, not a crutch.

We are entering a period in which AI-enablement – skills, workflows and governance - will be the prerequisite for successful AI integration. Or as a managing partner of a large national firm recently said: ‘A fool with a tool is still a fool.’ Law firms need to institutionalise continuous AI learning and development in their organisation. It is a foundation for changing their ways of working, while AI is being integrated.

5. Organise data and knowledge management with a view to AI

Generative AI’s potential is only as strong as the data it draws from. Clean, structured and accessible firm data is the foundation for effective use of AI. This means reviewing how knowledge is organised, whether client and matter data are properly governed, and how secure it is. ‘Garbage in, garbage out’ is not a cliché here: without robust data practices, firms will find it difficult to fully benefit from AI.

Law firms should invest (time) in the organisation of their knowledge management, also, because proprietary data is an important asset of law firms and valued by their clients. Being able to use internal knowledge effectively and efficiently increases the value of service to clients.

6. Embed governance and transparency into firm management

AI governance cannot sit in a standalone policy document. It must be part of the firm’s overall risk and compliance framework, with board-level oversight, including cybersecurity considerations specific to AI systems. This includes clear guidelines on when and how AI can be used, quality assurance processes, and client-facing transparency.

General counsel increasingly demand to know whether and how outside counsel uses AI (and in the future, where AI is not used), how quality assurance is measured and their data is protected. Law firms should therefore not only create an AI policy, but also monitor quality and compliance, address AI-related cybersecurity risks (including data breaches, model vulnerabilities and third-party vendor security) and adopt a client protocol. They should be able to explain how AI was used and how output was validated.

7. Working with integrated solutions

Isolated AI tools will not transform law firms; integrated systems will. Driven by law firm demand, legal AI solution providers are integrating with other systems which are relevant in legal practice, starting with legal data sources, document management systems and operating systems. Firms should expect this integration to continue with more systems in the coming year.

The combination of integrated systems and agentic AI will have a substantial impact on the way lawyers will work and the type of work they will do. This means law firms should focus on how to organise their data management and the transparency of the legal work process and the (legal) output their lawyers will deliver. It is important to take an integrated approach.

8. Anticipate client transformation and shifting demand

General counsel are adopting AI for purposes such as contract drafting and management, compliance monitoring, negotiations and risk analysis. The use of AI will change their role and the services they will request from outside counsel.

Law firms must start to anticipate how demand patterns will shift: clients will no longer need external support for routine tasks, but will value counsel who can co-develop workflows, add value as trusted advisors and provide judgment where AI ceases to work (eg, think about the nuances of stakeholder management, the impact of human interaction in negotiations, and the creativity in new types of legal questions). Clients do and will increasingly expect law firms to have a deep understanding of workflows which can be made more efficient by AI use and will increasingly demand that this be reflected into speed of execution and legal bills.  

Law firms should actively engage with clients on how they use AI, what their AI roadmap looks like and what they expect from their law firms in the medium term. Position the firm as a partner in the client’s AI journey, not a follower; clients have their own AI journey to develop and appreciate learning from their law firms. In addition, law firms should map (potential) changing demand and identify new service opportunities and use this as input for their client and practice strategy moving forward. Experience shows that many clients are open to active engagement on AI and appreciate proactive reach-out from their law firms.

9. Build the business case and rethink the business model

With the effects of AI on the efficiency of service delivery and the investment needed (eg, for solutions, training, change management) firms must build realistic business cases that capture both efficiency gains and new value creation. This is, like in any digital transformation, the biggest challenge law firms will face in the coming time: how to manage this transition. Success requires treating AI as a horizontal enabler integrated across all firm functions and aligned with firm-wide transformation goals, rather than as isolated technology projects.

Discussions with law firm clients like general counsel, show that such general counsel will expect that pricing models will change from billable hours to outcome-based and value-based pricing. Law firms should quickly learn to use AI to reduce cost in low-value work and invest freed capacity into higher-value services. Explore new pricing models that reflect both efficiency and enhanced client outcomes.

10. Redefine people development in the age of AI

Perhaps the most critical topic we see in many AI discussions in law firms, bar associations, universities and legal departments is: how do we develop junior lawyers when the tasks that once trained their judgment are automated by AI? If AI drafts first versions and automates legal processes, junior lawyers risk losing the friction that sharpens reasoning, argumentation and language. Law firms must deliberately design learning paths where junior lawyers still wrestle with complexity, comparing alternatives and justifying their choices. Promotion criteria must reward judgment and innovation, not throughput (billable hours) alone. In parallel, new roles will emerge – AI adoption specialists, AI auditors – that require fresh skills.

Law firms should get back to the drawing board and redesign professional development to ensure juniors build judgment, not just oversight skills. Combine AI-assisted work with deliberate ‘AI-off’ sprints, counter-anchoring exercises and critical annotation of AI output. Create new career paths that blend legal expertise with AI fluency.

In summary

With the AI developments going fast, not only within law firms, but also within its clients, law firm leaders should step up and start to take a strategic and transformational approach to AI. Learning and adapting fast is important. Law firm leaders should use the topics mentioned in this memo to start planning for the future. We hope you find these quick tips useful. If you do not, tell us why. If you have better ideas, pass them on. If you can use them to help others, do so. AI is not replacing lawyers, it is changing what clients expect lawyers to be. And those expectations are rising fast!