Aude Mutter, a lawyer at the Nice Bar, and Yves-Marie Le Bay, Doctor of AI and Professor at Université Côte d’Azur and EDHEC Business School, led a fascinating conference on the growing impact of AI in the legal profession. Their presentation addressed three major areas: the gradual integration of AI into legal practices, the ethical challenges it raises, and the evolution of lawyers’ roles in the face of this technological transformation.

Far from being merely a technical revolution, AI raises fundamental questions about the practice of law, lawyers’ responsibility, and the guarantee of professional confidentiality.

 

AI in Law: Adoption and Resistance

AI is no longer a novelty in the legal sector. According to a 2024 Wolters Kluwer study, 75% of legal professionals and 68% of lawyers already use generative AI at least once a week. However, its adoption remains gradual, due to technical challenges and professionals’ reservations.

Concrete Uses of AI in Law Firms

Legal professionals are gradually integrating AI into their practices, particularly for:

  • Legal monitoring: automatic updates of regulations and case law, sparing lawyers tedious research work.
  • Document drafting: production of contracts, letters, legal documents, with an automated first draft.
  • Administrative task automation: email management, appointment scheduling, or billing.
  • Case preparation assistance: case law summaries, identification of relevant precedents, and analysis of large files.

These tools allow for considerable time savings, especially for low-value tasks. A lawyer can thus devote more time to strategic case analysis and client relations.

Why Is Adoption Still Limited?

Despite these advantages, only 18% of lawyers report using AI in their daily legal practice. Several barriers explain this caution:

  • Mistrust of tool reliability: generative AI, while powerful, can produce errors or legal “hallucinations” (references to non-existent legal provisions).
  • Ethical and confidentiality issues: protecting client data is imperative for lawyers, who must ensure that the solutions they use respect professional confidentiality.
  • Tool costs: legal software integrating AI, such as Doctrine, Rodali, GenAI-L from Dalloz, JP Intelligence Lexbase, LexisNexis, represent a significant investment for firms, with annual subscriptions of several thousand euros. These costs are sometimes shared depending on the bar associations. ChatGPT, for example, being trained on American law, software specialized in French law is essential.

 

Between Generative AI and Predictive AI: What Use in Law?

Distinguishing Generative AI and Predictive AI

AI applied to law can be generative or predictive:

  • Generative AI (e.g., ChatGPT, Mistral, Gemini) produces text based on statistical language models. It is used for document drafting, text summarization, or communication assistance.
  • Predictive AI, on the other hand, relies on analyzing past data to anticipate future trends. For example, it can be used to predict the likely outcome of a trial, based on previous judicial decisions.

Towards Predictive Justice?

Predictive AI raises numerous questions about the evolution of the judicial system. In some Anglo-Saxon countries, it is already used to:

  • Assess the risk of recidivism of a convicted person (e.g., the Open-Door Assessment System tool in England).
  • Help magistrates rule on cases based on statistical trends.

In France, the use of predictive justice remains marginal. However, its development could have major implications for the profession. If AI becomes capable of accurately assessing the chances of success in a trial, alternative dispute resolution methods (mediation, arbitration) could be strengthened. Some fear a dehumanization of justice, where decisions would be dictated by algorithms rather than judges. But for lawyers, these tools could also enable them to better advise their clients by assessing the risks of each legal action.

 

Ethical and Legal Issues of AI in Law

The use of AI by lawyers cannot occur without oversight. Several major questions arise:

Lawyers’ Responsibility for AI Errors

If a lawyer relies on AI to draft a document and it contains a legal error, who is responsible? The professional who used it, the software publisher, or the algorithm itself? Currently, responsibility always lies with the lawyer. They must therefore verify and validate every piece of information provided by the AI. This duty of diligence is all the more essential as AI hallucinations are a reality. A famous case in the United States saw a lawyer present fictitious legal references generated by ChatGPT to the court, resulting in disciplinary sanctions.

Respecting Professional Confidentiality

Professional confidentiality is a fundamental pillar of the lawyer’s profession. However, using solutions like ChatGPT poses a major problem: where does the data processed by AI go? By default, queries sent to ChatGPT are stored on American servers. This poses a risk of sensitive information leakage and non-compliance with GDPR.

To address this problem, some AIs, like Mistral, offer solutions hosted in Europe. Additionally, agreements are underway to ensure that European firm data remains protected when they use tools developed by American companies.

Legal Framework for AI in Law

The legal profession is regulated to prevent the unauthorized practice of law. However, legal chatbots are beginning to appear in the United States, capable of providing automated advice. This raises a crucial question: can we allow an AI to provide legal advice to an individual without human intervention? In France, only lawyers can provide legal counsel. Any automated solution must therefore comply with this rule, or face penalties for unauthorized practice of the profession.

 

The Future of Law in the AI Era: Towards an Augmented Lawyer?

Facing these transformations, the lawyer’s profession will not disappear, but will evolve profoundly.

  • The augmented lawyer: a professional assisted by AI: it will not replace lawyers, but it will enable the most innovative to optimize their work. Those who adopt these tools will have a clear competitive advantage, particularly in terms of speed and efficiency.
  • An increasingly specialized profession: with the rise of AI, repetitive and generalist tasks will be automated. Lawyers will therefore need to specialize more to offer added value that AI cannot provide: in-depth case analysis, legal strategy, and human relationships with clients.
  • The need for continuous training: as AI evolves rapidly, lawyers will need to train continuously to master these tools and understand their limitations. Training in legal tech and AI ethics will become essential for effective practice.

AI is therefore profoundly transforming the practice of law. While it brings considerable productivity gains, it also raises major ethical and legal challenges. The lawyer’s profession will evolve toward the role of an augmented expert, capable of leveraging AI to better serve clients, while ensuring the rigor and ethics essential to the profession.