Share with your General Counsel / CLO
In-house legal teams at Thoughtworks, SAP, Teleperformance, and Zepto are declaring the pilot phase over, setting a 2027 target for AI embedded in production workflows rather than sandboxed experiments. The clearest signal from this panel at ETLegalWorld’s AI-Powered Legal Transformation Summit is that the organizing principle has flipped: start with a broken operational process, then design the AI workflow, not the other way around. Teleperformance’s 18-month roadmap, Zepto’s structured internal training program, and SAP’s cross-functional governance committee all point the same direction.
What this means for your business
The single most useful diagnostic in this piece is the repeat-usage test. If your team grants access to an AI tool and employees use it once out of curiosity and then stop, you haven’t adopted anything; you’ve run an expensive demo. Legal departments that will actually reach production scale by 2027 are the ones that can name a specific repeatable process, assign a human owner to the AI output, and build an escalation path when the model gets it wrong. That operational specificity is what separates the teams on this panel from the majority still circling.
The governance framing here is worth taking seriously, even though the summit venue gives the panelists an incentive toward optimistic timelines. The argument that governance must move at the speed of AI volume is correct and underappreciated. A single global policy framework with local procedural layers, the model Teleperformance is building across APAC’s fragmented regulatory landscape, is the only architecture that doesn’t collapse when you’re running 20 simultaneous AI workflows across five jurisdictions. The alternative, sequential legal review after each deployment, is already too slow for the cadence most businesses are setting.
The pricing shift Kalarikkal flags deserves a budget conversation now, not in 2027. Law firms investing in proprietary AI will not pass savings through as a blanket discount; they will rebundle the value as outcome-linked fees on high-volume, low-complexity work while defending hourly rates on complex matters. If your outside counsel spend is concentrated in the first category, the negotiation you’re walking into next renewal looks structurally different than the one you had two years ago. The legal teams building internal capacity for that same work are the ones with actual pricing leverage.
Concept deep-dive: Agentic dashboard
An agentic dashboard is a monitoring layer where AI doesn’t just surface data but acts on it without waiting for a human to issue instructions, think of it as a legal operations system that pages you only after it has already flagged an unpaid invoice or a contract deadline at risk. Teleperformance is building toward this. The business relevance is that it shifts legal from a reactive function to one running continuous surveillance across contracts, obligations, and counterparty behavior in near real time.
Based on reporting from Legal AI Adoption: Legal AI is moving beyond pilots as teams focus on workflows, skills and governance for 2027, ETLegalWorld, originally published 2026-08-07 05:11:00.
