Stinson LLP deploys Harvey firmwide for contract drafting, litigation prep, and M&A due diligence

Stinson LLP, a national firm with 16 offices and attorneys spanning regulatory, litigation, and transactional practice groups, faced a familiar productivity constraint: highly skilled attorneys spending disproportionate time on labor-intensive but routine tasks. Transcript review across voluminous testimony, clause-by-clause contract analysis, M&A due diligence summarization, and foundational legal research were consuming hours that could otherwise go toward client strategy and risk assessment. At a firm of Stinson's scale, this inefficiency compounds across hundreds of matters simultaneously — representing a meaningful drag on both capacity and client service quality.

Stinson conducted an internal evaluation before selecting Harvey, a domain-specific generative AI platform built for legal and professional services. More than 100 attorneys across multiple practice divisions participated in a structured pilot to validate use cases before firmwide commitment. The deployment integrates Harvey's large language model capabilities directly into attorney workflows for automated transcript review, deposition outline generation, standard contract clause analysis, initial M&A due diligence summarization, legal research acceleration, and litigation trend analysis. Stinson also deployed Ask LexisNexis within Harvey, embedding LexisNexis's authoritative legal content directly into the AI environment — grounding generative outputs in established legal resources rather than relying solely on model-native knowledge.

Pilot attorneys reported measurable gains across three dimensions: speed, consistency, and collaboration. Teams identified concrete ways the platform freed time previously spent on high-volume, lower-judgment tasks — enabling more hours directed toward analysis, strategy, and risk mitigation. Feedback was described by Managing Partner Allison Murdock as "overwhelmingly positive," with attorneys across regulatory, litigation, and transactional groups reporting meaningful workflow improvements. The strength of pilot results led directly to a firmwide rollout commitment, framed explicitly as a long-term strategic investment rather than a technology experiment.

• Pilot scope: 100+ attorneys across multiple practice divisions
• Use cases validated: transcript review, deposition outlines, contract analysis, M&A summarization, legal research, litigation trend analysis
• Outcome: firmwide adoption authorized following pilot

• A pilot with 100+ attorneys across diverse practice groups builds the internal evidence base needed to secure firmwide buy-in — anecdote alone is rarely enough.
• Pairing a generative AI platform with an established legal content provider (LexisNexis) addresses the hallucination risk that makes solo LLM use problematic in legal contexts.
• Framing AI adoption as a long-term operational investment — not a test — changes how attorneys engage with and trust the tooling.
• Responsible AI governance (data privacy, security oversight) should be built into the rollout structure from day one, not retrofitted after deployment.