AI and the legal field — Transforming how we work

For legal professions, the arrival of AI has been a game changer. In a sector characterized by numerous texts, variables, and subtleties, AI saves considerable time and reduces mental load.

This was the topic of our roundtable discussion at our January AI Breakfast, where we had the pleasure of welcoming:

• Gérard Haas — Lawyer specializing in digital law, data, AI, and intellectual property, Haas Avocats
• Florence BUISSON-VINCENT — Legal Director. AI Legal Development, UIMM
• Pauline De Certeau — Director of GenIA-L and Digital Products, Lefebvre Dalloz
• Camille Fagart — Head of Sales, Doctrine

To learn about their experiences and how they use AI in their daily professional lives.

“In a world where everything is moving faster and faster, it is reasonable to wonder whether lawyers will still wear robes in the future.”

What are the uses of AI that are truly transforming legal practices?

One of the recurring tasks in the legal sector remains documentary research: legal texts, regulations, etc. There are many documents and complexities, and finding the right information can therefore take a considerable amount of time.

Camille Fagart tells us that one of the most popular use cases is therefore legal research via a conversational interface.

The employee explains their problem and expectations, and the AI intuitively suggests one or more texts that may answer their questions, citing its sources, without spending hours on the task.

But AI doesn’t stop there. It can also provide analysis, drafting assistance, legal summaries, and more… in short, it can provide operational support to legal professionals.

Without compromising legal rigor and the reliability of information, how can we build truly useful AI?

Legal professions require precision, and AI must always source its answers so that each employee can verify the accuracy of the agent’s statements. The reliability of responses depends above all on the quality of the data provided to the AI and how it is structured.

Beyond technical performance, the AI agent must also act as a “bridge” between the expert and the client. It must be able to explain, or even simplify, some legal concepts for a non-specialist clientele. This is known as “legal design.”

Beyond high-performance AI, a needs assessment must first be conducted upstream in order to match the capabilities of the AI solution to the real needs of the organization, but also to comply with data protection and GDPR principles. Using an AI solution provider allows you to get support on these two decisive points when deploying the solution.

What are the main risks associated with using AI in a legal context, and how can they be anticipated?

Our four speakers are unanimous: the major risk lies in the absence of human verification. Legal advice remains an intellectual service that entails responsibility, so it is essential never to take AI’s assertions at face value (beware of “hallucinations” where AI invents facts).

The protection of advice given by legal professionals is also an issue. A client may be tempted to take a report with advice from their lawyer or attorney and give it to a “generalist and consumer” generative AI tool to summarize or explain it to them. Raising awareness among the general public and clients is therefore also essential in order to preserve data security.

One of the risks that still exists today is that of hallucinations, or when AI provides false information. It is therefore essential to have access to an AI assistant that knows when it does not know something and does not give an answer rather than inventing texts that do not exist.

How can a legal department integrate AI while ensuring security, compliance, and internal acceptability?

Integrating AI can be a challenging process. For Florence Buisson-Vincent, the priority is data control: data must be relevant, recognized by lawyers, and secure. Beyond reliability and security, data must be “usable” by AI, which again requires a lot of structuring and drafting work to ensure effective AI.

Work teams are not left out of the equation. For AI to be widely adopted, it must fit seamlessly into employees’ business processes and the tools they use. It must be “naturally” integrated into their daily lives, with a simple interface and few but truly useful features.

These two key elements go hand in hand: AI that is well integrated into workflows, but also effective and secure, has every chance of being adopted!

How will AI transform legal professions in the coming years?

According to Pauline De Certeau, no profession in the legal world will escape AI, and it is the analytical aspect (particularly thanks to agentics) that will revolutionize these professions, enabling them to solve more complex problems with an almost infinite network of information.

But to truly take advantage of the potential of AI, the challenge lies in training and having even sharper free will than we do today.

Gérard Haas tells us that in the next 10 years, 8 of the 10 biggest systemic risks will be due to either climate change or digital technology. And in a world where everything is moving faster and faster, we are entitled to wonder whether lawyers will still be wearing robes in the future.

If you had to give one piece of advice to a legal professional who is hesitant to get started with AI, what would it be?

Florence Buisson Vincent: “For a lawyer who is hesitant, I would say, test legal and professional AI, benchmark it, and you will see that you will find added value in AI. Test, test, test, and benchmark.”

Gérard Haas: “Go for it, go for it, go for it, and then if it works, hire the best lawyers and they will defend you.”

Pauline De Certeau: “You have to hurry up and take it slow!”

Camille Fagart: “It’s obvious that you have to go for it, but beyond benchmarking, you have to ask yourself what is most important for your own legal department/law firm. You have to define a set of specifications, because there are many players out there.”

Conclusion

While AI is undeniably reshaping the legal profession, it is not happening overnight. Its successful integration depends on three key factors: reliable data, a clear strategy, and constant critical thinking. Much more than a simple tool, AI is proving to be a performance lever, freeing lawyers from repetitive tasks so they can reconnect with what matters most: client relationships and advice.

Want to free up your legal teams with reliable and secure AI? Contact us.