General AI models fall short for legal work. Specialized solutions are essential.
Off-the-shelf AI models don't solve the core problems lawyers face. Max Junestrand, CEO of legal AI platform Legora, said the industry initially believed fine-tuning general models would work. It didn't.
"Fine-tuning doesn't really seem to work at least on the scale that we were operating," Junestrand said. The legal sector's complexity demands purpose-built applications layered on top of foundational models, not modifications to the models themselves.
Legora operates across 800 customers in over 50 markets and recently raised $550 million at a $5.55 billion valuation. Junestrand co-founded the company at 23 with no legal background and grew it from 40 to 400 employees worldwide.
Law firms adopt AI to compete in a low-differentiation market
The legal market has embraced AI faster than many expected. Law firms operate in what Junestrand calls "pretty low differentiation" - a space where one firm's services closely resemble another's.
AI changes that equation. When one firm uses AI to offer better services at lower prices, competitors feel the pressure immediately. Adoption spreads because it's now a competitive necessity, not a luxury.
Decades of software neglect created the opening
The legal sector was starved for good software. "The legal sector was so underserved with great software for such a long time that there was a lot of built up problems that we could easily solve with LLMs but they were really hard to solve pre-LLMs," Junestrand said.
Large language models addressed problems that existed for decades. That pent-up demand accelerated adoption across firms.
Legal AI products must outperform base models
Lawyers know what general models can do. A legal AI product that doesn't beat those baseline capabilities won't gain traction. Tech-savvy lawyers ask a straightforward question: why should they pay for this?
"If you showed up with a legal AI product it had to be better than the foundation models, otherwise they were just gonna say why are you deserving of my dollars," Junestrand said.
AI software companies operate differently than traditional firms
The structural differences matter. AI software companies must understand model capabilities deeply and translate that understanding into differentiated products. As models improve, the competitive advantage shifts.
Features that matter today may be irrelevant in six months. "As models got better your features may not matter in six months," Junestrand said. This forces constant adaptation.
Product quality sometimes means delaying revenue
Legora spent six months not selling because the company wasn't ready to onboard a thousand lawyers daily. That pause prioritized reliability over immediate growth.
Junestrand said success requires investment in product and engineering alongside a "culture of reliability first." Skipping this step creates problems that compound later.
The legal market's shift to AI is real and accelerating. Firms that understand the difference between general models and purpose-built solutions will compete more effectively. Those that don't will fall behind.
Learn more about AI for Legal and Generative AI and LLM to understand how these technologies apply to your practice.
Your membership also unlocks:
