Translating the Tools & Trends That Shape the Legal Profession
    Rising Above the Noise

    The Race to Fit the Firm

    Laura Kennedy & Chelsey Lambert 5 min read
    The Race to Fit the Firm

    The Race to Fit the Firm

    Tuesday, May 26, 2026

    Memorial Day was yesterday. We're grateful to the men and women who gave their lives in service, and mindful that the freedom they secured is what allows the rest of us to do our work at all. We hope the day gave you room to remember them, and time with the people you care about.

    That's why this edition lands on Tuesday. Here's the week that was.

    Last week the legal AI market stopped pretending the model was the product. Harvey spent three days at its own conference launching tools that have nothing to do with drafting quality and everything to do with the layers around it — adoption, institutional knowledge, contract workflow. iManage opened its knowledge graph to the wider AI ecosystem. And a survey landed that explains why all of this is happening at once. The vendors are no longer competing on whose AI is smartest. They're competing on whose AI fits the firm.

    If you read last week's edition, this is the same argument viewed from the vendor side. The connective tissue arrived. Now everyone is racing to own a piece of it.

    May 20–21, 2026: Harvey's Three-Day Forum

    Harvey used its Harvey Forum in New York to make three announcements in roughly forty-eight hours, and taken together they say more than any one of them does alone.

    On May 20, Harvey launched Command Center, a tool to help firms measure, govern, and optimize their AI adoption. It answers natural-language questions about how usage differs across practice groups, offices, and experience levels, and it draws on anonymized, aggregated data from more than 1,500 Harvey deployments so firms can benchmark against the field. Command Center was built with design partners including Haynes and Boone, Foley & Lardner, Clayton Utz, Rajah & Tann, and Dentsu. This is not a drafting tool. It's a management tool — the dashboard a knowledge-management or innovation leader uses to see whether the AI investment is actually landing.

    The same day, Harvey announced a partnership with DeepJudge, the Swiss institutional-intelligence platform founded by former Google researchers. The integration brings a firm's prior work, decisions, and precedent directly into Harvey's workflows, while respecting existing access permissions and ethical walls. Work done in Harvey feeds back into DeepJudge, so the firm's knowledge base compounds over time. Harvey CEO Winston Weinberg framed it as closing the gap between the expertise embedded in a firm's prior work and a legal team's ability to apply it consistently.

    Then on May 21, Harvey announced Contract Intelligence, a system built specifically for in-house legal teams. It uses contract agents to run first-pass reviews, apply playbooks, generate redlines, and escalate the issues that need an attorney. Playbooks update automatically as new contracts are signed. It carries a waitlist for early access, with general availability expected in Q3.

    The pattern across all three points one direction. Harvey is expanding past the workflow tools that made its name — drafting, research, due diligence — into the management layer that makes deployments governable and the knowledge layer that makes outputs firm-specific. As the underlying models become more of a commodity, the competition moves to the surrounding layers. That's the whole story of the week in one company.

    May 19, 2026: iManage Opens Its Knowledge Graph

    Two days before Harvey's contract announcement, iManage made the iManage MCP Server generally available, connecting its governed knowledge to the broader AI ecosystem through the Model Context Protocol.

    The significance is in the direction of travel. For two years, the legal AI conversation worried about firm knowledge being locked inside document management systems where AI couldn't reach it. The MCP server inverts that. A firm's governed content in iManage — with its permissions and security model intact — can now be reached by external AI tools through a standard protocol.

    It also confirms how fast MCP is becoming the default. Lexsoft made the same move the same week, announcing that its T3 knowledge management system is now fully accessible via MCP. NetDocuments built MCP access into the Legal Context Graph two weeks ago. The protocol is becoming the connective standard for legal AI faster than most firms have a policy for it, and the firms that built proprietary integration approaches now have a strategic decision to make.

    In Other News This Week

    LawDroid launched a free, open-source Claude AI plugin for civil legal aid, extending the connective-tissue pattern into the access-to-justice space rather than the enterprise market.

    Stilta, a Swedish startup automating patent litigation workflows, raised a $10.5 million seed round led by Andreessen Horowitz — a sign that investor appetite for narrow, practice-specific legal AI hasn't cooled.

    New York's Department of Financial Services issued guidance on frontier-AI risk, urging regulated firms to ensure their cybersecurity programs can promptly flag the weaknesses advanced models can exploit. It's a reminder that the regulators are now writing to the capability, not the category.

    The CTO Gap Zone™ of the Week: Proving It

    The survey that explains the rest of the week came from Litera, which released its State of Legal AI: Spring 2026 Market Sentiment Report. The headline number: 85% of law firms say clients are now driving their AI investment decisions. More than half — 51% — report that a client directly influenced an AI investment decision in the last twelve months. Only 15% describe AI investment as still entirely internally driven.

    That alone would be a notable shift. But the harder finding is the one underneath it: firms are struggling to demonstrate AI value, and ROI ranked as the least important factor driving their decisions. Read those two together. Clients are pushing firms to invest, firms are investing, and firms cannot yet prove the investment is working.

    This is the gap. Not a technology gap — the technology arrived, as last week's edition argued. It's a measurement gap. Firms are spending to satisfy client expectations they can't yet quantify, on tools whose value they can't yet demonstrate. That's an uncomfortable place to make six-figure platform decisions.

    It also explains why Harvey built Command Center. A tool that measures adoption is a tool that helps firms answer the client's next question, which is no longer "are you using AI" but "show me what it's doing." The vendors see the gap. Whether firms close it with discipline or just buy another dashboard is the open question.

    What This Means For Your Firm

    Last week the architecture arrived. This week the market organized itself around the layers of that architecture — and that maps directly onto ALPS™, the ten-layer framework Laura developed at Circle Management Group for what a firm's technology ecosystem is and should become. Harvey's Command Center is a management-layer play. Contract Intelligence is a workflow-layer play.

    The knowledge layer is where the pattern gets loud. The DeepJudge partnership and the iManage MCP server are both solving the same problem NetDocuments' Legal Context Graph addressed two weeks ago: getting a firm's institutional knowledge — its prior work, its precedent, its decisions — into AI in a form that's usable and keeps permissions intact. Three vendors, three approaches, one problem, inside two weeks. When the same layer draws that much vendor attention that fast, it's a signal that the knowledge layer is where the next round of firm decisions will be won or lost. Different vendors, different layers, same recognition: a capable AI model is only the starting point. What turns it into something a firm can actually rely on is everything built around it — the knowledge it can reach, the governance that contains it, the workflows it fits into.

    For a firm leader, this comes down to sequencing. The Litera data says clients will keep pushing, and the pressure isn't going away. But buying a tool for every layer is not a strategy — it's a budget line that will be hard to defend when a client asks what it returned. The firms that do well here will decide which layers matter most for their practice, in what order, and how they'll measure each one before they sign. That's strategic work, and it's exactly the work ALPS™ is built to structure: a coherent view of the whole ecosystem instead of a layer-by-layer scramble.

    The vendors spent last week competing for the layers. Your firm's job is to decide which ones it actually needs, and to be able to prove the answer when the client asks.

    Laura Kennedy and Chelsey Lambert are Co-CEOs and Co-Founders of Legal Tech Decoded. Rising Above the Noise lands every Monday at legaltechdecoded.com.

    The Daily Brief

    Every morning, the most important legal tech developments — curated and context-rich — delivered before 8 AM.

    Free. No spam. Unsubscribe any time. Read by 28,000+ legal professionals.