Why The Smartest Deal Teams Are Keeping AI Inside The Deal Room

InnovationWhy The Smartest Deal Teams Are Keeping AI Inside The Deal RoomByRusty Wiley,Forbes Councils Member.for Forbes Technology CouncilCOUNCIL POSTExpertise from Forbes Councils members, operated under license. Opinions expressed are those of the author. | Membership (fee-based)May 28, 2026, 08:00am EDTRusty Wiley is CEO of Datasite, a leading SaaS platform used by enterprises globally to execute complex, strategic projects. getty​The documents that decide a deal’s outcome (contracts, financials and disclosure schedules) often sit inside a virtual data room (VDR), protected by controls designed to minimize risk. But those same controls can make the information hard to analyze at speed. Meanwhile, many AI tools that could help are still positioned outside the deal room, requiring exports and work-arounds that introduce governance and chain-of-custody concerns.That tension of speed versus control is the defining challenge for AI in M&A right now. The teams that resolve it will move faster, surface risks earlier and bring more discipline to decisions under time pressure.​The Case For In-Workflow AIAI in M&A only works when it is designed for trust. The question isn’t whether models can summarize a contract; it’s whether a deal team can use AI on sensitive materials without losing control—no stray copies, no permission drift and no gaps in auditability. The risks are real: leakage, weak audit trails, hallucinated answers and unclear accountability. The right approach keeps humans in the loop and builds security, transparency and oversight into the workflow.The practical next step is connecting AI to where deal data already lives. Instead of downloading files and re-uploading them elsewhere, deal teams want approved assistants to operate inside the systems that hold the documents so access controls and audit logs stay intact. Emerging standards such as the model context protocol (MCP) aim to give AI assistants a consistent way to request permitted content for...