Last week, we woke up to news that had a direct impact on the legal sector. Anthropic announced new legal capabilities for Claude (its generative artificial intelligence model, comparable to ChatGPT), and the markets reacted almost immediately. Several publicly traded companies operating in the field of legal information and professional software, such as Thomson Reuters, RELX (LexisNexis), Wolters Kluwer, Pearson, Sage, and even the London Stock Exchange Group, recorded significant declines.

The risk of this type of news is not in the announcement itself, but in how the legal sector tends to interpret it. For many, it is yet another sign that we are facing another technological revolution. The problem is that this interpretation puts the focus back on the tool, when what is really at stake is the service model on which the legal business is organized.

Seen with a little more perspective, Anthropic’s move does not usher in a new era in law or herald the disappearance of legal work. What it does is accelerate a tension that already existed in the market and that for years could be managed without becoming explicit from a strategic point of view. Artificial intelligence is not transforming law; it is accelerating a division that has always been there between commodity services, process services, and experience services.

For years, technological advances in the legal sector were relatively orderly. First came automation tools applied to clearly standardizable tasks, such as document review or contract comparison. Later, artificial intelligence solutions specifically trained for the legal field appeared, such as Harvey, which increased individual lawyer productivity. In both cases, the technology improved what already existed, but did not force a rethinking of the service model.

What is beginning to change now is not so much the capacity for analysis, but the level at which the technology is applied. It is no longer just a matter of assisting professionals, but of executing complete processes and integrating them into real operational flows.

Here it is worth introducing an important distinction. For years, much of the legaltech ecosystem was built on what are known as wrappers (solutions that wrap third-party artificial intelligence models and add an interface or prompts, but do not control the entire process or take responsibility for the outcome). While access to the models was limited, this approach made economic sense. As models are integrated directly into business workflows, that middle layer becomes more difficult to justify.

From this perspective, Anthropic’s announcement is not relevant because of what it allows us to do, but because of where it is positioned. Anthropic is no longer limited to offering a model and is beginning to directly control the application and workflow layer, entering the space where a significant part of legal work is now behaving like a process.

Part of the market reaction can be explained precisely by this. It is not just a technological improvement, but a shift in the value chain. When the infrastructure provider decides to capture workflows directly, the assumptions on which many business models were built begin to come under strain at the same time.

That is why the impact is not limited to large traditional providers. It also affects private startups such as Harvey and Legora. Not because their products lose their usefulness, but because their value proposition now requires more than just analysis or interface. It requires operation, traceability, and a clear answer to the question of who validates, who decides, and who responds when the output becomes real action.

This pressure is also beginning to be felt within the firms themselves. Last week, Baker McKenzie confirmed a review of its global business services functions after analyzing its working methods, including the use of artificial intelligence and efficiency measures. In recent months, firms such as Clifford Chance, DWF, BCLP, and Freshfields have announced similar moves. It is not a question of replacing lawyers, but of redefining which work should be managed as a process and which work still requires expertise.

Some firms have opted for a different strategy. Instead of relying exclusively on external providers, they have decided to build their own capabilities. Garrigues is a good example. Its sustained commitment to innovation and digitalization responds to a long-term vision of how to integrate technology into its service model.

All of this helps to better organize the debate. At one extreme are commodity services, repeatable work with a low level of ambiguity, where the client seeks speed and price. In these services, the discussion is no longer legal but operational. The relevant indicators are no longer the sophistication of the analysis, but metrics such as response time, volume, or cost per unit. Technology does not eliminate these services, but it does compress their economic value.

At the center are process services, where clients seek consistency, coordination, and control over time. This is the layer where technology is having the greatest impact and where the most visible organizational adjustments are taking place.

At the other end are experience services. These are problems where information is incomplete, the context is specific, and the consequences of a bad decision are significant. Here, the lawyer is not just a provider of legal answers, but an advisor who understands the business and makes risk decisions alongside the client.

The impact of artificial intelligence is not only technological, it is strategic. And that is precisely why a reactive strategy does not work. It forces firms to clearly define which part of their work is commodity, which part should be managed as a process, and where their expertise really lies. But that definition cannot remain at the conceptual level. It requires internal organization, the creation of efficient workflows, and a rethinking of the operating model. Teams, capabilities, systems, and management criteria must respond to the type of service offered. When they do not, technology amplifies the inconsistencies of the model.

That is why the debate on artificial intelligence in law is not a discussion about tools. It is a discussion about business models and professional responsibility.

We are not facing a new technological revolution. We are facing an acceleration that drastically reduces the margin for strategic ambiguity. In an environment where technology executes processes faster and faster, not deciding what kind of firm you want to be ends up being, in itself, a decision.

Antonio Gómez Montoya – Partner at Black Swan Consultoría