
LexisNexis has announced the US customer preview of its new ‘Protégé General AI’ offering, which expands upon its existing agentic capabilities within Protégé.
So, what does this mean in practical terms? Here’s a few key points:
- ‘Protégé General AI offers a privacy-encrypted solution that eliminates the need to switch between tools.
- With a toggle you can engage with Advanced Reasoning and Deep Research for high-complexity legal issues or simple topic exploration and more.
- Protégé General AI also lets users choose the latest models, including GPT-5, GPT-4o, and o3 from OpenAI, and Claude Sonnet 4 from Anthropic, that they want.’
In short, think of it as a widened and more flexible version of what they had on offer before, giving the user a lot more customization, and easier ability to bring in the latest LLM’s reasoning capabilities.
There’s a bunch of other aspects, of which some of the key ones not already mentioned include:
- ‘A fully encrypted Lexis+ AI environment ensures data privacy protections that far exceed what consumer-grade AI tools offer.
- A Shepard’s Citation Agent: The advanced citation agent proactively checks any included legal sources for further human review – (AL Note: which is very handy to prevent hallucinations and errors).
- User and Org-level Control: Users can easily toggle Protégé General AI on or off, and firms can turn off General AI for users entirely to meet firm or regulatory policies.’
So, if you’ve played with GPT-4o, for example, and tapped its reasoning capabilities then you’ll have a good idea what’s possible here. Now imagine this operating within the LexisNexis legal data environment and on your desktop in a secure set up, and with the ability to use whichever LLM you want for each task. E.g. you want GPT-5 for X task, great, and when you want to jump over to Claude Sonnet, you can do that too.
This means you – the user – doesn’t have to keep engaging with lots of different AI companies yourself, you can get the benefits from them directly within the Lexis environment.
Overall, this brings you up-to-speed with what is on offer out there in the genAI world, but it’s fused together with what Lexis offers.
AL asked GPT-5 if this was a good move, here’s what it said:

And given that OpenAI is the provider of many of the foundational models used here, that’s useful to hear. (Of course, this is still an LLM talking here….so, maybe test Protégé out to know for sure.)
Sean Fitzpatrick, CEO of LexisNexis North America, UK, & Ireland, commented: ‘In response to customers’ requests for safe access to general-purpose models and greater control, we built Protégé General AI to put power directly in their hands, from selecting the model to guiding how it behaves in agentic workflows, all within a single, private environment.
‘Our goal is to support legal professionals with a personalized AI assistant that enables uninterrupted workflows and access to the full range of LexisNexis AI capabilities in one secure place. This vision is powered by the agentic AI innovation we’ve been advancing since 2024.’
—
P.S. Lexis has also kindly provided a breakdown of what is different between several models (and that’s useful as people who don’t spend all day testing LLMs may find it hard to immediately see the differences between them.)
Lexis noted the following:
- GPT-5 (Unified Reasoning): The latest model from OpenAI, it combines strong general knowledge and domain knowledge with reasoning and non-reasoning capabilities. LexisNexis is among the first to integrate this model through its direct collaboration with OpenAI.
- Claude Sonnet 4 (Natural Fluency): The latest model from Anthropic, it provides thoughtful and well-structured responses to support everyday tasks, communication, and brainstorming.
- GPT-4o (General Exploration): Excels at everyday tasks like general exploration, brainstorming, drafting, and integrating web-based information into real-time conversations.
- OpenAI o3 (Advanced Reasoning): Optimizes tasks like Deep Research, strategic decision-making, intricate problem solving, specialized workflows and nuanced understanding.
P.P.S. Approximately 200 law firms, corporate legal departments, and law schools are participating in the Protégé General AI U.S. Customer Preview Program. The preview begins in Q3 with general availability expected later this year.
—
You can find more about Protégé here.
—
Legal Innovators Conferences in New York and London – Both In November ’25
If you’d like to stay ahead of the legal AI curve….then come along to Legal Innovators New York, Nov 19 + 20, where the brightest minds will be sharing their insights on where we are now and where we are heading.

And also, Legal Innovators UK – Nov 4 + 5 + 6

Both events, as always, are organised by the awesome Cosmonauts team!
Please get in contact with them if you’d like to take part.
Discover more from Artificial Lawyer
Subscribe to get the latest posts sent to your email.