At the ETLegalWorld AI-Powered Legal Transformation Summit 2026, business and legal leaders argued that enterprise legal transformation will be defined not by how many AI tools organisations deploy, but by how deeply legal, compliance and governance get embedded into everyday operations.
The panel, “The Legal Operating System for Large Businesses: Where AI, Compliance and Governance Meet,” was moderated by Indranil Choudhury, CEO & Co-Founder of Lexplosion, with Alok Lall (Ex-COO, Microsoft), Dr. Bhaskar Ghosh (Ex-Chief Strategy and Innovation Officer), Sambit Das (Head of Legal, Porter) and Kishore Shetty (General Counsel, South Asia, GE HealthCare).
Das said the biggest danger in fast-moving businesses is decisions being made and implemented before legal implications are considered—often by frontline employees dealing with customers or regulators. Technology alone can’t fix this, he argued: organisations first need a culture where legal and regulatory awareness is instinctive. “First is building that culture itself, and then everything else follows,” he said. The goal isn’t legal approving every decision, but embedding compliance thinking into business decision-making itself.
Shetty added that legal exposure should be flagged when a business activity begins, not when a regulatory notice lands on the GC’s desk. If a company sets up a new factory, the system should surface licensing and regulatory obligations upfront. He compared this to car navigation: it can guide the route, but the driver remains responsible for the destination—AI can monitor risk, but human judgment stays central.
Toward an Event-Driven Legal Function
Ghosh argued legal problems rarely originate in the legal department, they’re triggered by business events like entering a new market, hiring a contractor, or launching an operation. “The future of the organisation is how do you create a more event-driven legal function,” he said. Rather than organising legal tech around silos like contracts or ethics, enterprises need systems that recognise a business event and connect all its legal, contractual, regulatory and ethical implications simultaneously, as they happen.Lall distinguished a specialist legal tool from a true enterprise operating system by where information originates and what it triggers. A real legal operating system must capture signals from across the business and translate each into the right obligation, owner, timeline and preserved evidence.
The key question, he said, isn’t whether AI can generate information, but whether the enterprise can build a reliable chain from signal to obligation to owner to action to evidence. AI can help construct decisions; it shouldn’t yet inherit them.
Das said centralised compliance platforms create accountability beyond the legal department by mapping obligations clearly so employees see what they own and when. He argued organisations should make non-compliance “expensive”, with technology improving discoverability and management visibility.
Shetty noted boards expect risk visibility across HR, manufacturing, finance and operations, but legal cannot independently know every operational risk. His solution: functional and factory heads certify obligations within their remit before board meetings, shifting responsibility closer to those managing the activity, reinforced by technology through reminders and escalation.
AI’s Real Value: Verification, Not Just Generation
Beyond chatbot-style tools, the panel highlighted AI’s role in verifying compliance evidence, checking uploaded documents actually match the correct entity, period or requirement, rather than just marking tasks complete. Choudhury noted AI can also catch blank or irrelevant uploads submitted merely to close a workflow.
Ghosh cautioned against adopting AI simply because “the train has already left”, the real question is “why AI?” He split AI’s value into automation (speed, efficiency) and intelligence (better judgment), insisting every initiative needs traceability, explainability and a human owner.
He also flagged change management as the bigger obstacle, resistance often comes not from leadership but from employees whose workflows are disrupted when AI compresses days of work into minutes.
Choudhury said a governable system must handle exceptions – unresponsive approvers, conflicting departmental data, tasks marked complete despite unresolved obligations, while preserving a defensible audit trail, which matters more than AI’s ability to draft content.
Beyond Volume, Toward Better Judgment
Lall argued AI’s real opportunity isn’t processing more contracts faster, but freeing lawyers to dig deeper, test assumptions and stress-test decisions. Choudhury agreed the benefit lies less in creating spare time than in tackling risks organisations currently can’t manage manually.
The panel concluded that the goal isn’t an AI-powered legal department, but an enterprise model where legal risk is caught earlier, compliance is embedded operationally, and AI strengthens—rather than replaces—human accountability.
The AI-Powered Legal Transformation Summit 2026 by ETLegalWorld is currently being held in Bengaluru. The summit has brought together general counsel, in-house legal leaders, and technology innovators, to chart the next phase of legal transformation in an AI-driven world.
