Advocate, advise, and accompany

These are the three essential roles lawyers will play in the post-AI era. We need to start preparing legal education, lawyer licensing, and law practices to adapt.

The range of possible outcomes from Generative AI’s entry into the legal sector covers a wide spectrum, with two straw men anchoring the extreme endpoints.

At one end is the Pure Skeptic, who says Gen AI will make no difference to how lawyers work and how legal help is delivered, since it will prove too error-prone and insufficiently specialized for practical use in the law. At the opposite end is the True Believer, who says Gen AI will become so powerful that it completely replaces lawyers, consigning them to the Island of Obsolete Professions alongside farriers and icemen.

Obviously, Gen AI’s actual effect on the law will land somewhere between these two extremes. But not all landing points on that spectrum are weighted equally. As we move towards the maximum-impact end, the risks to the legal sector as we know it don’t just rise. They increase exponentially. The impact curve goes: 📈.

Suppose the skeptics are mostly right: Gen AI will be useful in some respects but won’t be a game-changer for legal services. If so, then you can ignore the hype, save your money, and plan for a future that generally resembles the present. The risk here is overreaction, wasting valuable time and attention preparing for a storm that never arrives.

But suppose the believers are mostly right: Gen AI will trigger profound, even shattering changes in the legal landscape. If so, then legal institutions rooted in that landscape (law firms, law schools, courts) will face serious if not existential challenges to their structure and viability.

Even if you think this latter outcome very unlikely — and I’m not at all sure that it is — the payload of this outcome is incredibly high, because the impact would be so severe. A very tiny possibility multiplied by an extremely destructive event equals enough overall risk to constitute a legitimate concern. From a risk management point of view, therefore, we need to take seriously the prospect — no matter how small some might believe it to be — that Gen AI will bring transformative change to the legal sector.

Consider this scenario:

Ten years from now, Generative AI has proven capable of a stunning range of legal activities. Not only can it accurately write legal documents and conduct legal research and apply law to facts, it can reliably oversee legal document production, handle contract negotiations, monitor regulatory compliance, render legal opinions, and much more. Lawyers are no longer needed to carry out these previously billable tasks or even to double-check the AI’s performance. Tasks that once occupied 80% of lawyers’ billable time have been automated.

What are the chances this scenario unfolds within the next ten years? You can decide that likelihood for yourself, but I think anything above 1% represents the potential for major disruption to the legal profession. No lawyer or law firm could survive an 8% annual reduction in their inventory for ten straight years, or anything close to it. If the risk here is even 1 in 100, then we owe it to ourselves to make plans for this contingency.

I’ve been thinking hard about this possibility ever since ChatGPT-4 first entered the public domain 11 months ago. I’ve been asking myself (and I’ve had lawyers ask me): If Generative AI proves to be as capable of performing lawyer work as we imagine, what will lawyers still be needed for? I’ve posed this question before in the context of Gen AI, and I’ve taken a couple of shots at answers. Now here’s a fuller response, for your consideration.

Even in the foregoing scenario, even if Generative AI has swallowed most of our existing activities ten years from now, lawyers will still be needed. They will fulfill three essential roles that are completely beyond the capacity of AI and that will remain incredibly valuable to clients. Lawyers will advocate, advise, and accompany.

1. Advocate. A lawyer will personally represent a client (an individual, business, or organization) and truthfully advance or protect their interests before a legal arbiter or in the public square, using knowledge of the law and human society and the skills of reasoning, persuasion, and rhetoric. The lawyer will advocate on the client’s behalf, “standing in” for the client without personally joining to the client’s position. The lawyer will advocate within ethical boundaries in legal settings and social boundaries within public settings.

It’s a tired cliché that “no computer will ever get up in court to defend a client,” but it also happens to be true. Only humans can advocate. This isn’t just because argument and persuasion are human skills meant to move human feelings and decisions. It’s because the very act of advocacy, of standing up to say “I speak for this person,” is a personal commitment, one that signals to the audience and the community that the person and their position deserve to be heard. A computer could be programmed to argue a position. But only a human can make the meaningful personal commitment that advocacy represents and where much of its value resides.

2. Advise. A lawyer will provide a client with informed guidance and useful direction regarding difficult or important choices facing the client. The lawyer will gather relevant information about the client’s situation and recommend one or more courses of action that the lawyer believes will advance or protect the client’s interests. Using the human qualities of judgment, experience, wisdom, and character, the lawyer will identify options and suggest which one(s) seem best, but will never override the client’s own judgment or unduly influence the client’s eventual decision.

One of the oldest and most revered lawyer titles is “counsellor,” borne out in the act of advising a client. To advise is more than just “to inform” — clients don’t even need Gen AI to find out their legal entitlements online. To advise is to guide with assurance, counsel with confidence, offer direction with perspective. For years, clients facing hard choices have told their lawyers, “I know what the law says; but what do you think I should do?” In future, they’ll substitute “the AI” for “the law,” but the question will be the same. The you will be the same. People need good advice from trusted advisors, and that will not change.

3. Accompany. A lawyer will join a client on their journey, providing steady support and companionship, for as long as the client wants and for whatever the client might need. “Accompaniment” encompasses advice and advocacy, but it reaches beyond these transactional roles. To accompany someone is to make an ongoing personal commitment of presence, interest, and reliability — to affirm the person’s worth and their journey’s importance through the ongoing expression of personal engagement. A lawyer who accompanies a client offers them a relationship, for no purpose other than the relationship itself.

Advice and advocacy are centuries old in the legal profession; yet I think, for lawyers, accompaniment might be our essential service. I borrowed the term from the world of ministry — specifically, from a non-profit I support called Catholic Christian Outreach, which describes accompaniment as purposefully journeying alongside someone towards their spiritual goal. I think, in a secular sense, that also perfectly describes this role of a lawyer: to maintain an ongoing relationship of supportive presence, offering insights or conversation or even just quiet presence, while the client journeys towards their life or business objectives. It might be the most human thing a lawyer can do.

We keep talking, early in this AI era, about lawyers gearing up to do “higher-value” work. Is there anything higher-value than these three activities? Is there anything less susceptible to automation, even from 99th-percentile AI? Are there activities more meaningful to lawyers, more valuable to clients, more likely to lead to long-term retainers with healthy remuneration? Would that every lawyer everywhere was engaged in advocacy, advice, and accompaniment for clients every day.

But we’re not, of course. These moments are rare jewels in a typical lawyer’s life, a dazzling flash of real fulfillment before the dull routine of drafting and reviewing and phoning and transacting and docketing reasserts itself. Senior lawyers see these flashes more often; junior lawyers might never experience one before abandoning the profession in bitterness. Our highest-value activities are also our least often enjoyed.

Up until now. Generative AI, in the scenario above, poses an enormous threat to the way lawyers work and the lives lawyer lead. But it also presents a tremendous opportunity to exchange all our darkness and dread and drudgery for advocacy and advice and accompaniment. Or maybe it will be less of an opportunity and more of a mandate. Maybe, if Gen AI does become all that it might be, then lawyers will be forced to become all they might be.

This kind of future will require a massive overhaul of our legal habits and institutions. Aspiring lawyers will have to develop and demonstrate judgment, maturity, persuasion, empathy, insight, passion, character, perspective, wisdom, leadership, and a host of other skills on top of applied legal knowledge and abilities. Regulators will have to reconfigure entry-level competence around the proven ability to advocate, advise, and accompany. Lawyers and law firms will have to price their services around the delivery of these three kinds of services and the client’s experience in receiving each one.

Maybe there’s a 1% chance any of this comes about. Maybe it’s higher. Is that a risk you feel good about simply dismissing? Is it an opportunity you feel good about potentially missing?