8 out of 10 lawyers use it to prepare their case

All these advantages are present Jacinto Vicente, lawyer specializing in the field of sports. At 59, he has not yet taken the step of trying this type of tool, but he says he is thinking about it more and more, because customers are putting pressure on them. “It’s not a subject that I know much about, but I know they will demand it of me,” he admits.

Both Castizo and Vicente belong to a not-so-distant generation of lawyers who They spent weeks preparing their file and that, on many occasions, they had to rely on entire teams to process the documentation.

Today, this scenario is changing dramatically and, according to the European Legal Technology Association (ELTA), 80% of lawyers already use generative AI in their weekly routine and half do so daily. Nine out of ten people say technology significantly improves their professional performance.

What do the AIs they use look like?

“AI does not replace legal judgment, but rather augments it with more comprehensive data, precedent and analysis,” he says. Alexandre CastellanoCEO of Maite.ai, one of the many tools leading this revolution in the office.

Castellano explains that these systems combine different language models with gigantic legal documentary bases that let you locate items, compare resolutions, or detect patterns in seconds. In the specific case of Maite, he claims to have tested the tool by submitting it to the judicial access examination: the platform obtained 96 points out of 100, above the threshold of the last call for human applications.

When asked about the future of the profession and the risk of over-delegating to technology, Castellano says he believes every office will continue to retain his own identity. He explains that these assistants draw conclusions not only from laws and resolutions, but also from internal documentation that each office decides to upload to the platform. Contract templates, procedure templates or own clauses are now part of the AI ​​working environment. “If their contracts are better than those of the competition, that’s how they will appear to Maité,” he emphasizes.

An uncertain future for the profession

Knowing which sectors will be most affected by AI is a question that has been on the table for decades, but which only began to be taken seriously with the massive emergence of language models in December 2022. The unprecedented success of ChatGPT revealed in a few weeks a truth that is difficult to accept: artificial intelligence does not threaten manual professions as much as intellectual professions. The famous Moravec paradox.

Overnight, professions that had always been considered safe, such as those in creation, ceased to be so. But if there is one group that is truly exposed, it is those who filter and process large volumes of information, such as translators, analysts, consultants… and, of course, lawyers. But then, what future can we predict for justice?

“Brilliant but unstable intern”

Several tests suggest that AI will not eliminate this type of profession in the medium term, even if it will widen the distance between experts and novices. The Harvard Professor Karim Lakhani argues that this technology acts as a force multiplier for those who already dominate a field. According to his thesis, a lawyer with a decade of experience today revises and writes faster than ever, while a newcomer can barely control what the AI ​​gives him. The gap that is opening is not so much between humans and machines, but between those who can detect errors and those who depend on the correctness of the answer.

Ethan Mollickprofessor at Wharton, insists the same thing. He describes language models as a “brilliant but unstable intern” who can produce complete documents only if there is someone sane to supervise him. Without this control, the system becomes a factory of errors that are difficult to identify. In other words, artificial intelligence rewards senior workers, while representing a huge barrier to entry for newbies.

First effects on the labor market

This asymmetry is starting to be felt in the labor market, particularly at entry-level positions. A recent Stanford study detects a drop of nearly 13% in junior position openings in sectors most exposed to automation with AI – programming, customer service, administrative and accounting tasks – while vacancies for more senior profiles remain the same or even increase.

In the UK, another analysis reports a nearly 32% drop in entry-level jobs since the launch of ChatGPT, and voices like that of Dario Amodei, Anthropic’s CEO warns that up to half of junior administrative jobs could disappear within a few years as companies replace human hours with software agents.

In law firms, this movement is also starting to manifest itself. Thomson Reuters data shows that the proportion of junior associates to total billers is falling, while the middle and upper layers are reinforced. In fact, some companies are already openly questioning whether it makes sense to maintain large pyramids of associates when most of the repetitive work can be farmed out to software and a handful of associates oversee the outcome.

Disappearance of the hourly pricing model

Regardless, the risk of a generation of lawyers without middle levels is not the only change on the horizon. The very structure of the company also begins to weaken. The strategy expert Rita Gunther McGrathprofessor at Columbia Business School, said last Thursday in the Wall Street Journal that AI is pushing companies toward a model in which value trumps hours.

In her analysis, the expert ensures that the emergence of systems capable of revising thousands of contracts in a few minutes or drafting complex documents in a few seconds makes the temporal variable lose all its economic meaning. And above all, it shifts human contribution to areas that are not measured in hours, but in judgment, creativity and relationship management.

Added to this is the pressure from customers themselveswho no longer agree to pay for the training of junior profiles and are starting to demand prices adjusted to the real value delivered. “The future of professional services will not lie in the hours recorded, but in the quality of human input,” concludes the expert.