GO Public: Inside Cooley's ChatGPT Work-Built System for IPO Preparation
Cooley, a law firm advising on 180 global deals worth more than $51.5 billion in 2025, has built a proprietary AI product called GO Public on top of OpenAI's ChatGPT Work. The system is designed to help lawyers move faster through the mass of information that IPO preparation generates, while keeping human review at the center of every step that matters.
Cooley, a law firm advising on 180 global deals worth more than $51.5 billion in 2025, has built a proprietary AI product called GO Public on top of OpenAI's ChatGPT Work. The system is designed to help lawyers move faster through the mass of information that IPO preparation generates, while keeping human review at the center of every step that matters.
A firm with a specific claim to make about IPOs
Cooley did not build GO Public in the abstract. The firm has spent more than two decades advising on more venture-backed IPOs than any other law firm, and it has held the top spot in United States issuer-side IPO work for a decade. That track record is the backdrop against which GO Public was developed: not a general-purpose legal AI experiment, but a tool shaped by a firm that has handled a specific, high-volume category of transaction longer than almost anyone else. When a firm with that kind of concentrated experience decides to build its own AI product rather than buy an off-the-shelf tool, it says something about how particular the workflow is.
What the harness actually does
GO Public runs on an agentic harness, a structure that draws a hard line between what the system's agents can do on their own and what must go back to a lawyer for review or validation. The harness pulls together client information, public sources, and curated precedents, and analyzes that material so a lawyer isn't starting from a blank page. But the analysis is explicitly a starting point for review, not a substitute for it. The distinction matters because it defines the product's actual scope: GO Public organizes and surfaces information across a defined set of steps, and the harness is the mechanism that decides, step by step, where automation stops and legal judgment begins.
Who actually built it, and with whom
GO Public came out of work by Cooley's legal engineers, innovation counsel, and practicing lawyers, built on ChatGPT Work and developed in collaboration with OpenAI on legal subject-matter expertise and AI engineering. That combination is worth noting: the people shaping the product's legal logic are the same people who use it, not an outside vendor guessing at legal workflows in isolation. The collaboration with OpenAI supplies the underlying model and engineering support, while Cooley supplies the domain structure the harness is built around.
Where this could go next, and where it stands now
Cooley has stated that it sees potential for GO Public beyond IPOs, across capital markets transactions more broadly, though the firm has not detailed what that expansion would look like or when it might happen. For now, GO Public is a Cooley-built, Cooley-internal product: there is no separate signup path, pricing page, or public release for outside firms to adopt it directly. What it does offer is a concrete example of a large, established legal practice choosing to build its own agentic system on ChatGPT Work rather than treat AI adoption as a generic add-on. For firms or teams evaluating a similar path, that means the practical entry point isn't Cooley's tool itself but the same underlying platform and the same kind of structured collaboration between subject-matter practitioners and AI engineers that produced it.
Specs
- 2025 global deals: 180
- 2025 deal volume: more than $51.5 billion
- Primary use case: IPO preparation
- Underlying technology: OpenAI ChatGPT Work
Sources
- OpenAI — OpenAI