In recent months, I have met with partners eager to make up for lost time, convinced that their competitors have pulled ahead in the adoption of artificial intelligence. Some were held back by poor advice: build their own AI system, as if that were a straightforward undertaking. Others, faced with a legitimate concern such as protecting the confidentiality of their clients’ information, chose not to move at all.
In reality, their competitors have indeed gained an advantage—but not in the way they imagine. While many firms remain focused on deciding which AI tool to acquire, we have watched the conversation move on. In the firms that are truly leading, partners are no longer discussing the technology itself or what it can do. Instead, they are grappling with a far more significant challenge: how talent should now be evaluated.
Two years ago, I wrote in this newspaper about junior lawyers and artificial intelligence. At the time, I deliberately took a cautious position. I argued that AI would accelerate the technical development of junior associates, but not their judgment or their ability to generate business, and that this could ultimately lengthen the path to partnership. What I expressed cautiously then has, over time, become the starting point of a much broader challenge.
To understand why, it is necessary to separate two things that have always evolved together. Every lawyer develops along two learning curves. The first is the technical curve: drafting a contract, researching precedent, reviewing large volumes of documents in a transaction. The second is slower and less visible: developing the judgment to know what advice to give when the law is unclear, understanding what clients fear but never explicitly express, building the ability to bring work into the firm, and sustaining trusted relationships over many years.
Until very recently, these two curves progressed at roughly the same pace because they were nourished by the same source: time spent working on real matters. That is why seniority mattered—and not by itself. The traditional up-or-out model continuously filtered talent along the way, so anyone who remained at the firm after several years had demonstrated both technical competence and broader professional capability. A seventh-year lawyer had, almost by definition, accumulated more of both than a third-year lawyer; otherwise, they simply would not have made it that far.
Artificial intelligence has disrupted that balance. It dramatically accelerates the first learning curve while barely affecting the second. Today, a third-year associate can produce technical work that previously required several additional years of experience. They can review hundreds of comparable transactions in the time it once took to examine only a handful and produce a solid first draft within a single afternoon.
Yet that same associate still cannot exercise judgment, win clients, or guide them through complex decisions in the way a seventh-year lawyer can, because those dimensions of the profession do not develop faster simply because the tools improve. They continue to be built—and passed on—through repetition, observation, mentoring, personal mistakes, client meetings, and years of exposure that cannot be compressed.
What two years ago appeared to be primarily a training issue has now become something much larger. Added to that challenge is a new one: measurement. How should a lawyer now be evaluated? How should they be compared with their peers? On what basis should firms decide who deserves promotion, how much they should be paid, and—most importantly, in my view—which people deserve the firm’s investment?
The entire architecture of the modern law firm was built on the assumption that years of practice, more than any other variable, were sufficient to organize and assess talent. That assumption underpins the up-or-out ladder, determining when an associate is promoted or asked to leave; it supports the lockstep compensation model, common in many firms, where remuneration increases with seniority; and it even shapes how firms price the work of mid-level lawyers.
Once seniority no longer reliably predicts what someone is capable of doing, all of those mechanisms remain anchored to a metric that has ceased to be a fair and reliable measure of talent.
The challenge becomes apparent in everyday situations. A third-year associate delivers technically flawless work, thanks in part to AI, yet we know they are still not ready to handle a client independently. We believe a lawyer is prepared for the next step, only to discover that they know how to draft the document but not how to think about the client’s business—because they have not yet learned what it means to think like an entrepreneur. Or promotion season arrives, and we are faced with two associates who joined the firm in the same year and whose résumés are nearly identical. One produces work much faster with AI, while the other, although slower, is the one who truly understands clients and consistently brings in new business. Seniority, which once settled these questions almost automatically, no longer tells us which of the two should advance.
From this follows an important economic consequence. The easy assumption is that if AI lowers the cost of technical legal work, legal fees can only move in one direction: downward. The opposite is more likely to happen. The work that AI can generate becomes less valuable precisely because it is increasingly accessible to everyone. What retains—and indeed concentrates—its value is what AI cannot compress: interpretation, judgment, the ability to decide what to do with the information the technology produces, and the trust built with clients over many years. Pricing will inevitably shift toward those qualities. But before firms can charge for them, they must first identify who within the organization actually creates that value.
For Latin American law firms, the challenge has its own distinctive dimension. Across the region, many firms still operate with highly pyramidal structures in which seniority determines both authority and compensation, all at a time when demographic trends were already narrowing the base of the pyramid. Add to that a technology that blurs the very metric on which those decisions have traditionally rested, and the people responsible for evaluating talent—whether the managing partner or the firm’s partnership committee—find themselves deprived of the single most convenient tool they once relied upon.
Firm leaders now face a difficult task. Each firm must redefine, according to the kind of institution it aspires to be, what constitutes a good lawyer at every stage of a legal career, now that technical productivity and professional judgment no longer develop in tandem. That requires a different approach to evaluation, greater candor about who is ready to move forward and who is not, and a fundamental redesign of the firm’s systems for career progression and compensation—systems that, in many cases, remain calibrated primarily around seniority. This is a strategic decision because AI narrows the technical gap between the average lawyer and the exceptional one. As a result, what increasingly distinguishes lawyers is their judgment and the value they create for clients.
The question we are now forced to ask is one the profession has not seriously confronted for decades: by what measure do we distinguish talent now that simply counting years is no longer enough?
Seniority, by itself, never measured very much. What made it a reliable proxy was the up-or-out system, which gradually filtered out those who failed to meet the firm’s standards. Remaining at the firm became the proof. If someone was still there after several years, it meant they had earned their place, and the signal itself was objective—a number that no one questioned. Artificial intelligence disrupts that mechanism because it is now possible to achieve technical excellence without having developed the judgment that the traditional filter implicitly assumed. Finding a new way to distinguish those who merely produce work faster from those who also exercise sound judgment—without sliding into subjective impressions—is the challenge law firms must now solve.