Legal Engineering Definitions
Legal Technology Stack Comparison 2025: What Leading Firms Are Building
By Anthony Villanueva ·
The legal technology stack — the combination of platforms, tools, and infrastructure that supports AI-assisted legal work — looks significantly different in 2025 than it did just two years ago. This comparison covers the core layers and the leading solutions at each.
The legal technology stack has undergone a structural shift in the past two years, driven by the rapid maturation of large language model technology and the emergence of legal-specific AI platforms built on top of it. Understanding the current stack landscape requires thinking in layers: document management infrastructure, AI model layer, workflow and process automation, data and analytics, and practice-specific tools.
Role of document management as integration points
The document management layer remains dominated by iManage and NetDocuments for large law firms, with HighQ and SharePoint variants common in in-house environments. These platforms have become increasingly important as AI deployment points — both iManage and NetDocuments have developed AI integration capabilities that allow language models to access matter documents directly, and the security architecture of these platforms is a critical consideration for privilege and confidentiality compliance.
Consolidation of AI foundation models
The AI model layer has consolidated more rapidly than anticipated. OpenAI's GPT-4 and its successors dominate the underlying model layer for most legal AI applications, with Anthropic's Claude emerging as a strong alternative particularly for long-document tasks. Legal-specific platforms — Harvey, CoCounsel, LexisNexis Lexis+AI, Westlaw AI — use these foundation models as their inference layer while adding legal-specific fine-tuning, retrieval-augmented generation, and citation verification capabilities.
Fragmentation of the workflow automation layer
The workflow layer is the most fragmented and least mature. Most firms use a patchwork of automation tools — Zapier or Microsoft Power Automate for simple integrations, custom API connections for more complex workflows, and practice management platforms that have added workflow capabilities as secondary features.
The firms building proprietary AI infrastructure, exemplified by Kirkland's $500 million initiative, are investing heavily in this layer — building orchestration systems that coordinate multiple AI tools, data sources, and human review steps into coherent workflows. Wolters Kluwer's Future Ready Lawyer research identifies workflow integration as the primary capability gap in legal AI adoption.
Frequently asked questions
Which platforms continue to dominate the document management layer for large law firms?
The document management layer remains dominated by iManage and NetDocuments for large law firms, with HighQ and SharePoint variants common in in-house environments. These platforms have become increasingly important as AI deployment points — both iManage and NetDocuments have developed AI integration capabilities that allow language models to access matter documents directly, and the security architecture of these platforms is a critical consideration for privilege and confidentiality compliance.
Which two foundation models currently dominate the AI model layer for legal applications?
The AI model layer has consolidated more rapidly than anticipated. OpenAI's GPT-4 and its successors dominate the underlying model layer for most legal AI applications, with Anthropic's Claude emerging as a strong alternative particularly for long-document tasks. Legal-specific platforms — Harvey, CoCounsel, LexisNexis Lexis+AI, Westlaw AI — use these foundation models as their inference layer while adding legal-specific fine-tuning, retrieval-augmented generation, and citation verification capabilities.
What specific initiatives are large firms like Kirkland undertaking to address the workflow layer gap?
The workflow layer is the most fragmented and least mature. Most firms use a patchwork of automation tools — Zapier or Microsoft Power Automate for simple integrations, custom API connections for more complex workflows, and practice management platforms that have added workflow capabilities as secondary features.
The firms building proprietary AI infrastructure, exemplified by Kirkland's $500 million initiative, are investing heavily in this layer — building orchestration systems that coordinate multiple AI tools, data sources, and human review steps into coherent workflows. Wolters Kluwer's Future Ready Lawyer research identifies workflow integration as the primary capability gap in legal AI adoption.