Gilbert + Tobin scales ChatGPT Enterprise and Codex
OpenAI's own customer story frames the rollout around CEO-led commitment, formal governance and human accountability — one source only.
Illustrative image: an empty law-firm library at dusk, with a long timber table, glass partitions and shelves of bound volumes.
In a customer story published by OpenAI, Gilbert + Tobin is described as scaling ChatGPT Enterprise and Codex through three things rather than a procurement decision: CEO-led commitment, strict governance, and a named human who answers for every output.
At a glance
- The source is a customer story published by OpenAI itself, not independent reporting.
- Products named: ChatGPT Enterprise and Codex, in use at the law firm Gilbert + Tobin.
- Three building blocks named: CEO-led commitment, rigorous governance, human accountability for outputs.
- The source page was unreachable during editing (HTTP 403); this piece rests on the public preview text.
- Only one independent domain backs this story — our two-domain sourcing rule is not met.
OpenAI's customer story presents Gilbert + Tobin's use of ChatGPT Enterprise and Codex as a governance project rather than a software purchase: the chief executive backs it, formal rules bound it, and a person remains accountable for what the systems produce.
What the story claims
The piece sits in OpenAI's customer-story section. It names two products — ChatGPT Enterprise for everyday work and Codex for software tasks. It attributes the rollout to three elements: leadership commitment from the top, rigorous governance, and human accountability for results.
Nothing further is available to us: not how many staff use the tools, not since when, and not which tasks they cover. For that reason this article carries no adoption, time-saving or cost figures.
Why governance matters more in a law firm
Legal work runs on professional duties — confidentiality, conflict checks, liability for the advice given. A language model can carry none of them. A firm deploying generative systems broadly therefore has to keep a person identifiable at every step who owns the result. That is the emphasis OpenAI's account puts front and center.
What we could not verify
The source page did not load while this article was being written; the request returned HTTP 403. The basis here is the publicly visible preview text, not the full article. Anything that exists only in the full text is deliberately absent.
There is also the provenance question. This is vendor material about the vendor's own customer. Such texts are evidence of what a supplier says, not of what a deployment measurably delivers. No second, independent newsroom has examined the case as far as we can tell.
What to take from it
For firms and in-house legal teams, the notable claim is not the product choice but the sequence: ownership and rules first, rollout second. Whether it played out at Gilbert + Tobin the way it is described cannot be checked against the material we have.
FAQ
Which OpenAI products does Gilbert + Tobin use?
ChatGPT Enterprise and Codex are the two named. The text available to us lists no other products and no usage figures.
Is the source independent?
No. It was published by OpenAI about its own customer, and we found no independent review of the case.
How much time does the firm save with AI?
We have no reliable figure. We deliberately publish no time or cost savings here because we cannot trace one to a source.
