The AI use cases powering daily legal work | Harvey
How legal teams are using Harvey across top use cases identified in the Briefing AI Leaders Community Index.
The legal industry is no longer asking whether generative AI will transform daily work, but how quickly that transformation will take hold, and where it will deliver real value first.
Recent data from the Briefing AI Leaders Community Index, which surveys more than 70 senior leaders across leading UK law firms, paints a picture of a profession in transition. Adoption is clearly underway: 50% of firms have already deployed legal-specific AI tools, with use cases spanning the full lifecycle of legal work — from summarisation, data extraction, and large-scale document review to due diligence, research, drafting, and knowledge management.
Yet the same data reveals a gap between experimentation and true operational change. Only around 20% of firms say AI is embedded in standard workflows. That gap reflects the deeper challenge of turning promising tools into repeatable, day-to-day workflows.
Throughout the rest of this post, we explore what that looks like in practice, spotlighting how our customers are using Harvey across five top use cases identified in the Briefing survey.
Summarisation: from weeks of review to same-day insight
AI-powered summarisation is increasingly becoming part of how legal teams handle high-volume review and client delivery — enabling faster turnaround while maintaining consistency in output. By reducing the time spent on synthesis, it allows lawyers to focus more on higher-value analysis and strategy, reflecting a broader shift in how legal work is structured.
At Honigman, using Harvey for summarisation offers a clear example of how this plays out in practice. With Harvey, teams can process large volumes of complex material (particularly deposition transcripts) and turn them into structured, usable outputs in a fraction of the time.
In one client engagement, the firm delivered a full set of deposition summaries along with thematic analysis in less than a day, work that would have previously required several associates working for two weeks. Just as importantly, the workflow is designed to be iterative: As new deposition transcripts become available, the team can quickly incorporate them into the existing analysis without restarting the process.
Data extraction: turning documents into structured data
For Haynes Boone, AI has helped the firm evolve data extraction from a time-intensive manual process into a scalable, repeatable workflow embedded in daily work. Using Harvey’s Agent Builder, the innovation team creates structured Workflow agents that extract key data points across documents and organize results into formats attorneys can review and refine.
In one recurring project where attorneys need to extract and analyse data across a large set of documents, what previously required months of manual review can now be completed in a matter of hours. Harvey handles the first pass of extraction and structuring, allowing attorneys to focus on validating outputs and drawing insights rather than combing through documents.
“[With Harvey,] what once took months can now be completed in a matter of hours, freeing up attorneys to focus on the higher-value parts of the work.” Tony Capecci, director of practice innovation at Haynes Boone.
In the firm’s finance practice, the team designed a Workflow agent that extracts specific data points from documents and compiles them into a checklist. Harvey produces the first draft automatically, which then goes to a partner for review and approval. This approach helps shift AI-powered data extraction from a one-off task into a consistent, repeatable part of how legal teams operate.
Large-scale document review: continuous review of shared datasets
King & Wood Mallesons is redefining large-scale document review as a continuous, collaborative process rather than a one-time exercise. Using Harvey Vault, teams centralize large volumes of documents, like lengthy steering committee meeting minutes, into a shared workspace that supports ongoing review and analysis.
Instead of manually sifting through documents for each new task, lawyers can query the full dataset and generate structured review tables. This approach not only reduces the need for repetitive document-by-document review, but also creates a persistent resource the team can return to as matters evolve.
The same dataset can be reused across multiple workflows — from identifying which witnesses attended specific meetings to supporting affidavit drafting and building chronologies. By embedding Harvey into this process, document review becomes a more dynamic, shared foundation for analysis. As chief innovation officer Michelle Mahoney put it: “Not only does this save the legal team time, it also creates a ready-made resource they can tap into for any future tasks.”
Due diligence: rapid insight across data rooms
Due diligence is often defined by time pressure and the need to quickly understand large, complex data rooms. Using Harvey, teams at Hengeler Mueller can quickly generate overviews of documents, identifying key materials and surfacing potential issues more efficiently than traditional manual review.
Pierre Zickert, counsel and head of legal technology, shared: “Harvey helps us gain a rapid overview of large data rooms, allowing teams to identify key documents and issues significantly more efficiently than has ever been possible by human effort.”
This capability is accessible across experience levels. At Hengeler Mueller, even trainees just days into the firm are building sophisticated Workflow agents to support due diligence and statutory analysis, not simple prompts, but structured logic that can be confidently applied in real client work.
Drafting: accelerating the first draft
Drafting is often where legal work slows down, not because of complexity, but because getting started can be the hardest step. At Winston & Strawn, Harvey is helping teams move past that friction by making it easier to generate an initial draft and iterate quickly toward a final product.
Rather than searching for the perfect precedent or template, attorneys use Harvey to produce a strong first version and refine it through successive iterations. This shift allows lawyers to react to the substance of a document and improve it in context, rather than getting stuck at the outset.
“Harvey turbocharges our attorneys’ ability to take that first step so they can figure out and execute the next one.” Monet Fauntleroy, managing director of practice client services and innovation at Winston & Strawn.
As a result, drafting becomes a more fluid, continuous process — one that fits more naturally into the pace of daily legal work. By reducing the barrier to starting and accelerating iteration, Harvey enables attorneys to move more efficiently from initial idea to polished output, while maintaining control over the final result.
The examples in this post point to a broader shift — from isolated AI use cases to workflows that define how daily legal work gets done. As firms move beyond experimentation, the focus is increasingly on embedding these workflows into everyday practice.
