The legal tech gold rush just turned into a bloodbath for everyone not named Harvey or Legora.

The Summary

  • Legal tech funding hit $4.6B in 2025, but $3.8B in 2026 so far — Harvey and Legora took 51% this year versus 23% last year
  • Strip out the two giants: everyone else fell from $3.59B to $1.89B, a 47% crater with six weeks left in the year
  • Winner-take-most dynamics are hitting legal AI faster than any adjacent category, signaling that building agents for professionals isn't a land grab anymore

The Signal

Legal tech just gave us the clearest early read on how agent economics actually play out when the category matures. The gap between Harvey and Legora versus everyone else isn't just wide. It's structural.

Harvey and Legora raised $1.95 billion combined in 2026, capturing 51% of all legal tech funding. The other hundred-plus startups fighting for crumbs split $1.89 billion. That's down 47% from the $3.59 billion those same non-giant companies raised in all of 2025. With six weeks left in the year, that gap isn't closing.

This isn't a temporary flight to quality. It's what happens when investors figure out that professional services AI has distribution moats, not just product moats. Harvey has partnerships with top-tier law firms like Davis Polk and Latham & Watkins. Legora has enterprise contracts that take 18 months to close and another 12 to fully deploy.

"The concentration of capital is hard to miss — and the drop-off hard to ignore."

Once you're embedded in a law firm's workflow, the switching costs aren't about features anymore. They're about retraining partners who bill $1,500 an hour and don't have time to learn a new interface. They're about compliance audits and client confidentiality agreements and IT security reviews that take quarters, not weeks.

The implication for the broader agent economy:

  • First-mover advantage matters again. Not because the tech is hard to replicate, but because professional workflows are hard to infiltrate twice.
  • Horizontal plays are dead. "AI for lawyers" worked in 2024. "AI for M&A due diligence at AmLaw 100 firms" is what gets funded in 2026.
  • Anthropic and OpenAI aren't the only platform winners. The companies that verticalize agents for high-value knowledge work are becoming platforms themselves.

The Implication

If you're building agents for professionals, the window to establish distribution is closing faster than the window to build a better model. Legal tech is six months ahead of medical AI, accounting AI, and consulting AI. Watch what happens here.

For investors, the data is clear: either back the category leader with enough capital to become infrastructure, or find a micro-vertical so specific that Harvey can't be bothered. The middle is gone.

Sources

Business Insider Tech