.How law firm Fidal is redefining legal practice with AI
Corporate legal pioneer Fidal partnered with Sopra Steria Next to design a powerful generative AI tool. The aim is not to replace lawyers, but to augment their expertise and reinvent legal process.
Legal clients increasingly demand faster turnaround times, more precise answers and legal services tailored to a complex regulatory landscape.
As the legal industry evolves under the pressures of digitalisation and rising client expectations, Fidal – a French leader in business law for more than a century – recognised the need to modernise its tools while safeguarding its intellectual rigor.
AI to empower lawyers
To address these challenges, Fidal launched a digital transformation project to optimise the way it delivers legal services. The project led to the development of Fidal IA: an AI-powered assistant designed not to replace lawyers, but to augment their work by supporting research, drafting and legal analysis.
The firm’s ambition was to leverage artificial intelligence to streamline repetitive tasks, improve access to high-quality legal data and empower lawyers to spend more time on strategic thinking and client interaction.
Importantly, the solution had to respect the constraints of confidentiality, legal precision and internal governance. To bring this vision to life, Fidal partnered with Sopra Steria Next, the group’s consulting firm specialising in change management and digital transformation empowered by NExtGen (AI Cloud Data).
Fidal IA was developed through an iterative process beginning with a proof of concept tested by a dozen lawyers, followed by a phased rollout to over 1,250 of its 1,900 legal professionals. The system integrates a tailored AI model that draws on a vast and evolving base of legal documents.
Features include document summarisation, translation, reformulation and intelligent drafting using validated templates. A key innovation is the prompt library, which offers legal-specific prompts to guide effective use of the tool across practice areas.
Designing an AI solution for the legal profession came with multiple challenges. First, the volume and quality of legal data had to be carefully curated. Fidal processes over 500,000 legal documents each year, but not all of these are relevant for AI training.
Duplicates, outdated drafts and incomplete records were filtered out to build a trusted and anonymised internal database. This database was then supplemented with high-quality open data sources, including content from Légifrance and the BOFiP.
Empowering legal minds through
Sources: checked
Another major hurdle was cultural adoption. Lawyers are trained to exercise discretion and skepticism. As such, any AI solution needed to be transparent, reliable and supportive – not intrusive.
The firm approached this by involving lawyers directly in the project’s design, from identifying relevant use cases to validating outputs during the pilot phase. Special attention was given to the user interface, ensuring that the AI clearly cites its sources. Sopra Steria's use of retrieval-augmented generation technology enabled this by ensuring that the AI responses were grounded in verifiable documents, thus preserving the trust and reliability expected in legal advice.
Importantly, Fidal IA maintains the firm's high intellectual standards. Every response is reviewed and validated by lawyers, maintaining full professional accountability while improving responsiveness and efficiency.
The verdict is in
The collaboration between Fidal’s internal digital team and Sopra Steria Next was fundamental to success. Sopra Steria brought in AI and data governance expertise, while Fidal ensured that every functionality reflected the nuances of legal practice.
This collaborative governance model not only improved adoption rates but also ensured AI’s relevance to daily legal work. The solution is designed to evolve with continuous input from users and real-time updates as new legal content becomes available.
The implementation of Fidal IA has led to measurable time savings in legal operations. Tasks that once took hours – such as jurisprudence research, regulatory tracking or generating legal drafts – can now be completed in minutes.
On average, research time has decreased significantly. This efficiency gain is especially impactful for junior lawyers, who now experience faster learning curves and reduced reliance on senior colleagues. For experienced lawyers, AI becomes a strategic partner, freeing time for higher-value advisory work.
In an industry known for billable hours, this AI-enabled productivity tool is even leading Fidal to explore alternative flat-rate billing or subscription-based services – potentially offering clients greater transparency and value.