Agentic diligence goes live, Claude-connected practice ops reshape legal work, and junior roles shift
The gist
Legal work is shifting from AI-assisted drafting to AI-executed workflows, forcing lawyers to supervise agents, verify outputs, and redesign how teams review and delegate.
This week’s developments
Ropes & Gray Puts Agentic Diligence Into Production
Ropes & Gray and OpenAI launched a proprietary agentic M&A diligence tool that mirrors the firm’s deal-review methodology and produces issue-level reports in hours, not days or weeks, with findings tied back to source documents and final judgment left to lawyers. That is a meaningful step beyond the workflow systems covered last week: the output is no longer just extracted clauses or summarized contracts, but supervised deal work product that flags acquisition risks and questions such as notice or consent requirements.
The workflow layer is moving with it. iManage, Litify, Epiq, Reveal, and Google Cloud expanded AI-driven legal workflows across knowledge management, eDiscovery, and orchestration, while Gemini Enterprise for Legal stressed citation verification and said firms including Cleary Gottlieb, Freshfields, Weil, and Williams & Connolly are using it. Harvey’s new funding for Canadian expansion shows capital is backing geographic scale, not just pilots. Indian court guidance and draft regulatory materials also hardened the baseline: AI output is advisory, must be independently verified, and should not stand alone as evidence without disclosure.
For lawyers and legal ops teams, the job is now moving from assembling first-pass diligence to supervising AI issue spotting, checking source support, and documenting verification. The career edge is increasingly in making AI-assisted work traceable, defensible, and safe to rely on.
How should we adapt diligence staffing and training now?
If you're an individual contributor
- First-pass diligence is commoditizing; your edge is AI judgment.
- Learn to verify citations, spot gaps, and write defensible issue notes — that’s what keeps you indispensable.
Sources
- The AI due diligence dilemma — Lexpert, September 9, 2026
Shows where AI diligence helps, where it overwhelms, and how lawyers keep outputs accurate and defensible.
- AI Is Rewiring M&A Practice — But Judgment Still Sets the Deal’s Direction — The National Law Review, August 19, 2026
Practical guidance on checking AI outputs, framing diligence questions, and preserving lawyer judgment in deal work.
- 5 AI-Powered Due Diligence Strategies Every PE Firm Should Know — Analytics Insight, August 9, 2026
Practical strategies for using AI to screen contracts, risks, and alternative data while preserving expert review.
If you manage a team
- Your team’s value is shifting from review volume to supervised judgment.
- Coach lawyers to check source support and exceptions, and reallocate time from manual review to QA and escalation.
Sources
- How Young Lawyers Can Build A Career That Outlasts The Next AI Update — Above the Law, September 10, 2026
How young lawyers can build judgment, verify AI outputs, and develop reliable workflows that strengthen their careers.
- Patrick Forquer on Legora’s Agentic AI, Legal Engineering, and Consumption-Based Pricing — The Geek In Review, August 17, 2026
How legal teams overcome governance, permissions, and change management to use AI beyond basic document summarization.
- Microsoft releases new AI playbook for enterprises with real-world examples, and it reveals a surprising 'moat' you may already have — VentureBeat, September 17, 2026
Framework for mapping processes, setting human-agent boundaries, and building evaluation and governance layers before deploying agents.
If you lead the organization
- Manual diligence capacity is no longer the moat; governed AI is.
- Invest in traceable workflows, verification standards, and AI-literate talent before competitors reset speed and cost.
Sources
- When AI Agents Act, Where Should the Human Check Go? | Law.com — Law.com, September 24, 2026
Explains where to place human oversight in agentic legal workflows to preserve accountability and compliance.
- ILTACON 2026: Who's leading the flock? Balancing agentic AI automation and human oversight | Thomson Reuters Institute — Thomson Reuters, September 14, 2026
Framework for classifying AI tasks, enforcing audit trails, and balancing autonomy with human oversight.
- <b>AI is pushing CLM beyond contract repositories to actionable legal intelligence: Industry Experts</b> — ETLegalWorld.com, August 7, 2026
How leaders should structure CLM AI with data discipline, risk parameters, and human oversight.
8am Turns MyCase Into a Claude-Connected Operations Layer
8am’s expansion of MyCase with 8am IQ, including IQ Firm Agent and IQ Draft, plus MyCase MCP, is the clearest sign yet that the practice-management layer is being wired for direct AI execution. The MCP connector is the key signal: 8am says it is the first official Claude Model Context Protocol connector for legal practice management, letting users query and execute MyCase tasks from Claude across 73 actions spanning cases, clients, documents, billing, and intake. That cuts down the back-and-forth between systems for routine work.
8am also launched 8am Capital, a funding program tied to incoming payments, adding a separate financial layer to the platform. Taken together, the release suggests MyCase is becoming less of a record-keeping tool and more of a place where AI can draft, route, and complete operational tasks.
For legal teams, the practical shift builds on the governed execution trend already underway: more day-to-day MyCase work can now start in Claude and finish inside the platform, but supervision still matters. Firms should expect faster execution on routine matters, not autonomous judgment, and should keep validating outputs and actions before they affect clients, billing, or deadlines.
How should we adapt roles as Claude handles MyCase tasks?
If you're an individual contributor
- Routine MyCase work is moving to Claude; your edge is review and judgment.
- Learn to verify AI-drafted tasks, catch errors fast, and own the exceptions that still need human judgment.
Sources
- LawToolBox MCP for Microsoft 365 & Claude Connects Legal AI to Matter Data Containers in M365 — LawNext, August 24, 2026
Shows how LawToolBox MCP connects Claude to deadlines, pleadings, and matter workflows inside Microsoft 365.
- The Human Clipboard: Closing the Operational Gap in Legal AI — Artificial Lawyer, September 24, 2026
Shows how MCP connects assistants to case systems for auditable task execution with human approval.
- Why Systems Thinkers Are Better at Using AI — The AI Maker, September 8, 2026
Shows how to document tasks into reusable AI skills, routines, and agents for more consistent execution.
If you manage a team
- Your team’s admin load is shrinking; coaching quality control is the new job.
- Shift training from process steps to AI supervision, exception handling, and when to stop automated actions.
Sources
- Ari Kaplan — Pioneers and Pathfinders, August 6, 2026
How law firms can coach teams, redesign workflows, and scale small AI wins into lasting adoption.
- Meet the Leader: Why AI 'workslop' is a leadership problem — Radio Davos, August 20, 2026
Practical AI coaching and role-play for frontline managers handling feedback, supervision, and higher-stakes conversations.
- Steal My AI Marketing System (One Prompt Builds It) — Marketing Against the Grain, August 18, 2026
Shows how to build an AI coach that reviews usage and suggests better prompting and workflow habits.
If you lead the organization
- MyCase is becoming an AI execution layer, not just a system of record.
- Rework operating models and hiring around governed AI execution; invest in controls before routine work moves faster than oversight.
Sources
- Inside Track - From the field: How agentic AI is reshaping adoption at Microsoft — Microsoft, September 24, 2026
Microsoft’s framework for adoption, governance, and change management as AI agents take on completed work.
- Introducing The Collective — Supply Chain Now, August 3, 2026
Framework for executive buy-in, guardrails, and ROI-focused AI investment without blind adoption.
- What 20 Years of Software Investing Says About AI | Matt Hedberg — Run the Numbers, September 7, 2026
Framework for making bold AI bets, avoiding pilot sprawl, and turning AI into production value.