Legal AI Start-Ups Face Valuation Test

Legal technology start-ups are turning artificial intelligence into specialist products for law firms and corporate legal teams, but the model is beginning to face a sharper commercial test. Companies such as Legora are betting that frontier AI groups will leave enough room for industry-specific intermediaries to build valuable businesses around their models.

The proposition is simple. Rather than asking lawyers to work directly with general-purpose AI systems, these start-ups wrap foundation models from companies such as OpenAI or Anthropic inside legal workflows, training layers and user interfaces. The result is software aimed at routine legal tasks including document review, research and report drafting.

That packaging has helped drive rapid growth. Legora has recorded seven consecutive quarters of annual recurring revenue growth above 50% and has expanded through acquisitions, including London-based litigation technology specialist Wexler. Its pursuit of a possible $10bn valuation reflects investor enthusiasm for AI companies that can convert broad technological disruption into recurring industry revenue.

Yet the risks are substantial. A valuation at that level would imply a steep multiple of recurring revenue, leaving little room for slower growth or weaker margins. The economics may also become more difficult if companies continue to absorb token costs through subscription pricing, especially as more advanced agentic systems move towards consumption-based models.

Competition could come from two directions. Large law firms and corporate clients may decide to build proprietary AI systems, particularly where they have the capital, data and incentive to control their own tools. At the same time, the AI giants whose models sit beneath these legal products are moving closer to the sector themselves.

The opportunity remains real, but less protected than early growth suggests. Legal AI wrappers may have expertise, client relationships and workflow knowledge, but their long-term value depends on whether they can become indispensable platforms rather than temporary layers between lawyers and the model-makers.