AI in Law Firms: From Denial to Real Adoption, in Three Phases According to Olivier Chadut

Paris-based consultant Olivier Chadut, founder of an AI-native consulting firm, has identified three distinct stages in the legal sector's relationship with artificial intelligence. His diagnosis is blunt: most law firms are only now emerging from the second phase, and are finally doing so in earnest.

Three stages: from denial to action

The first stage was predictable: lawyers dismissed AI as inapplicable to their profession. The arguments went roughly like this: "Law is too complex and nuanced," "the client pays for our unique expertise, not for a machine," "a court will not accept a document written by an algorithm." This phase lasted longer than in most other professions, partly because of the profession's high self-regard and partly because of real regulatory risks.

The second stage revealed the nature of the industry more candidly. Organizations bought LLM licenses en masse, not for real work but for display. Conferences filled with polished slides about "AI transformation," and press releases announced partnerships with OpenAI or Microsoft. Inside the firms, a few enthusiasts experimented with ChatGPT while everyone else worked as before. The licenses were paid for out of marketing budgets, as a signal to partners and clients rather than for results.

The third, mature phase is arriving now. Its markers are concrete ROI metrics, specialized tools for specific tasks, and growing pressure from clients who are already aware of what AI can do.

Where AI already works for real

Real use in the legal industry today is concentrated in a few clearly defined areas:

• Legal research: finding precedents and analyzing case law (Harvey.ai, Casetext, LexisNexis+)
• Contract review: automatically flagging risky clauses and comparing against templates
• M&A due diligence: processing thousands of pages of documents in hours, not weeks
• Drafting: first versions of standard agreements, NDAs and employment contracts
• Billing and time tracking: classifying working time by matter without manual entry

The key change in the economics: partners used to see AI as a threat to the billing model, since fewer hours per task meant a smaller bill for the client. Now the logic has flipped. Freed-up time is redirected to complex strategic work, while routine tasks are repriced at competitive fixed rates.

Clients have become the catalyst

Large corporate clients, including banks, investment funds and industrial holdings, now ask their legal advisers directly: "Which AI tools do you use on our matters?" This pressure from below proves more effective than any internal initiative. Compliance departments and D&O insurers are also starting to impose new requirements on the speed and documentation of work.

At the same time, the market for tools is consolidating. Instead of dozens of unconnected pilots, large firms are choosing 2-3 specialized Legal AI platforms and embedding them in real workflows. The era of experiments is ending; the era of operational scaling is beginning.

What this means

The legal industry has come a long way, from categorical denial through showcase purchases to genuine automation. For practicing lawyers, this is not a threat of extinction but a shift in priorities: less routine research and draft work, more strategic expertise. For Legal AI companies, the moment of truth has come: the market has matured, clients are ready to pay, and the question is no longer "is AI needed in law" but "which of the competitors will win."