Key Takeaways

  • Morgan Stanley manages cap tables for nine of the top 10 unicorns, 75% of the top 100 by market value, and 50% of roughly 1,000 unicorns.
  • The firm deployed Debrief, an AI tool that listens to client calls and executes back-office operational tasks.
  • Advisers now act as approval checkpoints, clicking "agree" on background tasks like updating beneficiaries or opening trust accounts.
  • Head of Wealth Management Jed Finn argues the binary legal boundary between private and public markets should become a spectrum of disclosure tiers.

Turning Advisers Into One-Click Approvers

Most enterprise AI tools stall at summarization. Companies give staff tools to record calls and summarize meetings, but the actual administrative follow-up remains manual. An adviser finishes a call, reads the recap, and spends the next hour entering data across compliance forms and custodial systems.

Morgan Stanley is taking a different approach with Debrief, an agent tool for its wealth management team. Instead of drafting recaps, the tool listens to intent and executes operational workflows directly in back-office software.

Finn outlines the architecture: “the next iteration of that is an agent is listening and actually taking the action so if you say I want to uh change a beneficiary or I want to open up a trust account it will actually do that all in the background now and the adviser just hits agree” (play from 33:36)

This design changes the economics of service work. The human remains legally responsible and retains final sign-off, but their role shrinks from manual form entry to a single click. When an agent executes the pipeline directly, an adviser can handle far higher client volume without expanding back-office headcount. For builders creating internal software, that is the goal: do not generate text for users to copy and paste. Run the workflow in the background until it hits an approval gate.

The Cap Table Moat and the Private Market Spectrum

Morgan Stanley holds a massive position in private tech equity administration. Through its corporate platforms, the bank tracks the shareholder registries for the world's largest venture-backed businesses.

Finn points to their reach: “So we're the cap table manager now for nine of the top 10 unicorns.” (play from 29:37) Across the entire market, that presence holds firm: “Uh by market value for 75% of the top 100 and for 50% of all almost a thousand unicorns.” (play from 29:41)

Holding equity data for half of all tech unicorns gives Morgan Stanley direct insight into liquidity strains. Startups are delaying initial public offerings for years, raising private capital at massive valuations while employees sit on illiquid shares. Meanwhile, institutional and retail investors are clamoring for direct equity stakes in early AI leaders. As Finn notes: “There's a lot of demand. There's a lot of demand. I mean it it's tough, right? Because the TAM for intelligence is kind of all the dollars.” (play from 30:37)

The core problem is regulatory rigidity. Securities rules treat market access as a toggle: companies are either closed private firms or heavily regulated public entities.

Finn argues that framework is out of date: “So the first thing that I think is clear is that the difference between private and public ought not to be a binary distinction, right? What it ought to be is levels of disclosure and a gradation of that and then levels of investor sophistication, right?” (play from 34:41)

A graded system would let late-stage companies grant employee liquidity and tap broader capital pools by offering audited financials, without enduring the full administrative drag of public filings.

What to Do With This

Audit your team's workflow for tasks where an employee manually copies decisions from a client call into internal databases. Reconfigure those pipelines so an agent drafts the changes automatically, leaving your team with a single action: click approve.