Mistral AI Takes Aim at Legal Sector Through Expanded Harvey AI Partnership
Mistral will gain access to more than 1,500 customers that use Harvey across over 60 countries
Mistral AI is bringing its models to the legal sector under a partnership with the startup Harvey AI, taking aim at a lucrative industry where rivals like Anthropic are expanding aggressively.
The French artificial-intelligence company will feature on a list of models that San Francisco-based Harvey offers on its platform to help law firms and in-house legal teams streamline work in areas like contract analysis, due diligence, compliance and litigation. Mistral models will initially be available to a few customers based in the European Union ahead of a wider rollout.
Lawyers handle troves of text and data, making the legal industry fertile ground for large language models that can help sift through documents. The potential to automate tasks and give lawyers extra time to take on more cases is an incentive to pay top-tier prices for AI tools.
Mistral will gain access to more than 1,500 customers that use Harvey across over 60 countries, putting it on steadier footing as it jockeys for position in an industry where Anthropic has already made strides: The Claude maker recently added more legal tools known as plug-ins that paying customers can use through its Cowork assistant.
Mistral’s partnership with Harvey isn’t new—the companies announced a collaboration in May 2024 to work out how to support clients. In a joint interview, Mistral Chief Revenue Officer Marjorie Janiewicz and Harvey Chief Operating Officer Katie Burke said the two startups dipped their toes in the water at the time and were jumping in now that Mistral models had evolved.
“The difference here was before we were thinking about it as a pilot,” Burke said. “Now, this is a permanent partnership that we’ll make publicly available to Harvey customers.”
For Mistral, the move represents an opportunity to delve into a niche and highly technical arena where users expect increasingly capable models that understand jurisdictional nuances and comply with data-privacy requirements.
Janiewicz said Mistral had set its sights on financial services and other industries with strong data-protection needs since the early days of the startup, and that it made perfect sense to focus on legal as it also ticks the box around the need for security. Mistral could explore further partnerships to expand elsewhere, she added.
“You should definitely expect us to become more and more present in any industry that, as I mentioned, would really value control of the data and also customization,” Janiewicz said.
For Harvey, the partnership means it can offer more choice to clients by adding Mistral to its repertoire of models that includes Anthropic, ChatGPT maker OpenAI and Alphabet’s Google. Burke said over half of Harvey’s customers were outside the U.S. and it was critical for the platform to provide a model that understands a global context and works across multiple languages.
AI tools are improving at a dizzying pace, raising fears that the technology could wipe out jobs across legal, accounting and other industries where software is essential. Earlier this year, Anthropic’s release of new capabilities to automate processes like contract reviews and legal briefings triggered a selloff in software-as-a-service stocks.