Share with your General Counsel / CLO
Legal departments at Vodafone, Cognizant, Indegene, and several other enterprises are moving past AI pilots toward something harder: restructuring how legal work is actually organized. At the ETLegalWorld AI-Powered Legal Transformation Summit 2026, general counsel from across industries converged on a consistent position: AI adoption in legal is an operating-model problem, not a software procurement problem. The throughline across every company represented was that governance, data quality, and workforce change management have to be built before scale, not after.
What this means for your business
The companies at this summit that are furthest along share one structural trait: they treated AI adoption as a workforce redesign with a technology component, not the reverse. Dilip Manepalli’s “use-case factory” framing at Vodafone is a good diagnostic here. If your legal team hasn’t decomposed contracts, IP, and litigation into discrete sub-processes and mapped which ones are automatable, you aren’t behind on AI tools, you’re behind on the prerequisite work that makes AI tools useful. Legal departments that skipped that step and bought software anyway are now running expensive pilots that never graduate.
The governance structure that emerged across multiple panelists deserves attention. Hitachi’s three-layer model (users, approvers, enablers) and Cognizant’s dedicated responsible AI function sitting alongside legal, IT, and security point toward the same conclusion: no single function can own enterprise AI accountability, but diffuse ownership produces the same failure as concentrated ownership. The companies getting traction have a named accountable layer, usually legal or a joint legal-risk committee, with IT as the implementation arm. What Syngene’s Sahana Chandrika described as “privacy and ethics by design” is the operationalization of that structure, building regulatory requirements into the AI system at architecture time rather than bolting on compliance reviews after deployment. For any legal team with cross-border data exposure, waiting for Indian domestic AI regulation to mature before introducing safeguards is a genuine risk, not a scheduling option.
The ROI framing Manepalli outlined, prioritizing use cases by business value and ease of implementation before moving through proof-of-value to minimum viable product to scale, is the right sequence, and most legal departments are stuck at proof-of-value because they never defined what “value” meant before the pilot started. Blupine Energy’s contract standardization story is a useful benchmark: the win wasn’t speed alone, it was consistency and reduced dependency on tribal knowledge about where prior contract versions lived. If your legal team can’t articulate a similarly concrete before-and-after for at least one process, you don’t have an AI strategy, you have AI enthusiasm. I’d revise this assessment if any of these organizations publishes hard productivity or cost figures in the next twelve months, but the absence of numbers in this summary is itself a signal about where the industry actually is.
Based on reporting from Transforming Legal Departments with AI: Governance, Skills, and ROI Insights, ETLegalWorld, originally published 2026-08-07 03:20:00.
