Recent law firm moves around Claude, legal agents, and custom model work suggest a clear direction for professional-services firms: the value is moving from broad chat tools to specific workflows that match how teams actually work.
The market is moving toward workflow-specific AI
Hanson Bridgett's firm-wide rollout of Claude for attorneys and professional staff shows how quickly general AI tools can become part of day-to-day legal work when they are tied to real tasks.
The reported use cases are practical: summarizing deposition testimony and lengthy records, drafting routine correspondence and memos, comparing document versions, supporting due diligence review, and helping professional staff in operations, marketing, HR, finance, and knowledge management.
AI agents are becoming the useful unit of work
Claude for Legal now includes more than 90 named agents, each built around an end-to-end workflow such as Vendor Agreement Reviewer, DSAR Responder, Termination Reviewer, and Claim Chart Builder.
That matters for firms because it suggests a shift away from asking one general model to do everything. Instead, firms can build or adapt workflow agents around repeatable tasks, then tune the underlying skill, practice profile, and connectors to how the team works.
Custom AI may become a firm capability, not just a vendor choice
Kirkland's hiring and infrastructure work point to a possible move toward fine-tuning open-source models and building a more tailored internal legal AI platform rather than relying only on off-the-shelf tools.
For professional-services firms, that raises a strategic question: which work should stay on standard platforms, and which high-volume or high-value workflows deserve custom treatment because they are central to the firm's delivery model?
What this means for law and accounting leaders
The common thread across these developments is operational discipline. The firms that benefit most from AI will likely be the ones that define specific use cases, connect them to existing systems, and establish a clear review process.
For law and accounting firms, that means starting with tasks that are repetitive, document-heavy, and easy to measure. It also means paying attention to sources, jurisdictional context, and internal controls before expanding AI across the firm.
- Start with one high-volume workflow, not a general AI rollout.
- Treat named agents and custom workflows as operational tools, not novelty features.
- Decide early where your firm wants off-the-shelf AI and where it needs custom AI.
- Build governance around review, sources, and task-specific control before scaling.
Sources watched
- AmLaw 200 Firm Hanson Bridgett Goes All-In with Claude (Artificial Lawyer)
- Claude For Legal Has Over 90 AI Agents (Artificial Lawyer)
- Kirkland Hints It Could Fine-Tune LLMs For Own Legal AI Model (Artificial Lawyer)
