A Top-Five Law Firm Just Refused to Pick a Legal AI Vendor — So It Bought Both

On October 16, two of the best-funded companies in legal AI put out competing press releases about the same client on the same day. Harvey announced that Hogan Lovells Cadwalader had adopted its platform across three practice groups. Legora announced that the same firm was rolling its platform out globally, with more than 1,300 lawyers across 18 practice teams already using it on live matters. Neither announcement was wrong. The firm had signed with both.

That is unusual. Over the past two years, Harvey and Legora have built their businesses by winning exclusive, firmwide contracts at prestige firms — Latham & Watkins and Gowling WLG for Harvey, Herbert Smith Freehills Kramer for Legora, and dozens of others on each side. Law firm technology committees have generally treated the choice as binary: pick a platform, standardize on it, and avoid running two overlapping, expensive AI systems that do roughly the same thing. Hogan Lovells Cadwalader, the roughly 3,300-lawyer firm formed earlier this year by the merger of Hogan Lovells and Wall Street's Cadwalader, Wickersham & Taft, decided that logic didn't apply to it.

What was actually announced

According to Harvey's own announcement, the firm deployed Harvey across Corporate & Finance, Litigation & Arbitration, and Regulatory/IP, with lawyers building custom agents for knowledge management, capital markets work, and international arbitration. Legora's announcement describes a parallel, larger-footprint rollout: access across the Americas, EMEA, and APAC, with a pilot phase that already ran more than 1,300 lawyers across capital markets, banking and finance, M&A, disputes, real estate, and tax and regulatory work, out of offices including New York, Washington, London, Frankfurt, Paris, and Luxembourg. The firm also agreed to serve as a design partner shaping two unreleased Legora features — an agentic legal-research tool and a regulatory 'Monitors' product — testing early versions on live client questions before general release.

Both vendor statements quote the same person, CEO Miguel Zaldivar, delivering slightly different versions of the same argument. To Harvey, he said the firm was 'perfecting the ways to capture and apply the expertise' spread across its practices. To Legora, he said the platform was helping lawyers while 'maintaining the professional judgment that remains at the heart of what we do.' Independent reporting from legal-industry outlet Non-Billable frames it more directly, citing Zaldivar's line that 'there is no single AI solution that will meet every need of a modern global law firm.'

The governance layer is the real story

What makes this more than a procurement footnote is the structure the firm built around it. Non-Billable's reporting, corroborated by Unite.AI, describes a new AI Board tasked with deciding, deal by deal, whether a given need gets built in-house through ELTEMATE — the firm's roughly 100-person internal legal-tech unit, co-run by Sebastian Lach and Manon Cordewener — or bought from an outside vendor. Three new leadership roles were created alongside it: Paul Otto as managing partner for legal technology and AI, Oliver Campbell as chief transformation officer, and Bill Ferreira running a new commercial unit covering pricing and legal project management.

That is a materially different posture from firms that have simply flipped a switch on one vendor's platform. It treats Harvey and Legora not as a single strategic bet but as two inputs into a build-or-buy decision the firm expects to keep revisiting — implicitly betting that the legal AI market is not yet settled enough to lock into one supplier, even at the cost of running two procurement relationships, two training programs, and two sets of security reviews in parallel.

There is no single AI solution that will meet every need of a modern global law firm.

The case against hedging

The skeptical read is that dual-sourcing is expensive theater as much as it is strategy. Running two firmwide AI platforms means double the licensing spend, double the integration work with document management and billing systems, and double the burden on partners already being asked to learn one new tool, let alone two with overlapping functions. Industry reporting on the broader legal AI market notes that Harvey and Legora have increasingly collided head-to-head for the same large-firm contracts precisely because most firms conclude that consolidating on one vendor is operationally simpler — which is why dual adoption at this scale, rather than a short side-by-side bake-off, remains rare among the firm's peers.

There is also a practical test the firm hasn't yet faced publicly: what happens when Harvey's agents and Legora's research tools produce different answers to the same question on a live deal, and which one a partner is supposed to trust. Neither vendor's announcement addresses conflict resolution between the platforms, and the firm's own AI Board — built to arbitrate build-versus-buy decisions — doesn't obviously have a mandate to arbitrate platform-versus-platform disagreements on substantive legal analysis.

Still, the timing is notable. The deal lands three months after Hogan Lovells and Cadwalader formally merged into a roughly $3.6 billion-revenue firm, a combination that forced the new entity to rationalize overlapping technology stacks from day one anyway. Seen that way, running two AI vendors may be less a bold strategic stance than a byproduct of a merger that handed the firm two legacy AI relationships it hasn't yet decided to unwind — with the AI Board's real job being to find out, over the next year, whether keeping both was ever actually necessary.