Lexsoft links T3 legal knowledge tool to AI platforms
Mon, 18th May 2026 (Today)
Lexsoft has made its T3 legal knowledge management system accessible through the Model Context Protocol and introduced a Microsoft-based OpenAI vectorised indexer in T3.
The changes are intended to connect T3 with a wider range of artificial intelligence tools used by law firms and corporate legal teams.
T3 previously operated as a standalone knowledge management product. Through MCP, organisations can now connect it to MCP-compatible AI orchestrators and external legal AI platforms for knowledge search, retrieval and classification.
This allows legal teams to use T3 as a source of curated, reviewed knowledge documents within broader AI-driven workflows. Lexsoft says the system is designed to provide the exact contextual reference for any datapoint or text extracted through AI.
MCP access
T3 can now work with legal AI orchestrators including Microsoft Copilot, Claude and Gemini, as well as third-party platforms such as Harvey. The move reflects a broader push by legal technology suppliers to make knowledge repositories easier to use within generative AI tools rather than as separate systems.
For law firms and in-house legal departments, the main challenge is often not access to documents but access to reviewed material that can be trusted in day-to-day work. Lexsoft is positioning T3 as a managed source of legal knowledge rather than a route into broader stores of unreviewed files held in document management systems.
The new indexer is intended to change how users search that material. Traditional indexing typically relies on exact words or close variations, while vector-based search is designed to identify related concepts and interpret terms in context.
Lexsoft gave the example of recognising that "contract" and "agreement" may be closely related, while distinguishing between terms spelt the same but carrying different meanings depending on context. In practice, that could help lawyers retrieve more relevant precedents, clauses or know-how documents when phrasing varies across teams and jurisdictions.
Search changes
The new vectorised indexer is hosted within a customer's own OpenAI tenant in the Microsoft environment. According to Lexsoft, this helps organisations meet security and data residency requirements while retaining control over indexed information.
Lexsoft also says it does not index customer data itself. Customers can keep their existing indexer or move to the new version, allowing firms to choose whether to adopt semantic search immediately or continue with their current set-up.
Lexsoft operates in the legal sector across Europe, the United States and Latin America, supplying software and business process services to law firms and corporate legal departments. Its broader product offering includes document management, knowledge management, practice management and customer relationship management.
The update comes as legal technology providers and large software groups compete to make AI tools more usable in professional services settings, where oversight, document provenance and auditability remain important. In that context, systems that can connect knowledge stores to AI assistants without moving data outside approved environments have drawn increasing attention from firms assessing how to deploy generative AI in legal work.
Lexsoft framed the changes around the role of human-reviewed knowledge in AI-supported legal workflows. It said T3 is intended to sit between raw document collections and user-facing AI tools, giving lawyers access to approved material within the tools they already use.
"By combining MCP functionality, advanced semantic search, and human-centered review, T3 now plays a critical role in supporting the current wave of AI-led transformation of working practices in the legal sector," said Carlos García-Egocheaga, Chief Executive of Lexsoft.
"We are developing T3 so that the solution enables organizations to optimally use AI for knowledge management while ensuring safeguards through close human involvement and oversight. The best part is that through generative AI tools such as Copilot, Claude, and Gemini, or Harvey, T3 knowledge management is seamlessly embedded within the legal workflow. In effect, the solution becomes invisible. Lawyers can simply trust that they are accessing the best knowledge documents as a matter of course," he said.