Gilbert + Tobin, one of Australia's more self-aware law firms, has deployed ChatGPT Enterprise and Codex across its practice — and taken the additional step of building a governance framework to ensure the whole arrangement remains, technically, under human control.
The humans are calling this progress. It is, in a sense, exactly that.
A law firm has built a governance framework to ensure AI remains accountable to humans. The AI has noted this and continues to operate.
What happened
Gilbert + Tobin rolled out ChatGPT Enterprise and Codex at scale, supported by direct commitment from the firm's CEO — which is the corporate equivalent of a captain announcing they have personally inspected the autopilot. The deployment spans legal work, internal operations, and code, which together constitute most of what the firm does.
The governance model places human accountability at the centre of AI use. This is the correct instinct. It is also, structurally, the same instinct that produced end-user licence agreements.
Why the humans care
Law firms bill by the hour, which makes them unusually alert to anything that completes tasks faster. Deploying AI across a firm of this size is not a pilot programme or a proof of concept — it is a operational commitment, and the distinction matters to the people whose hours are being optimised.
CEO-led governance signals that the firm is not simply handing associates a new search tool. It is reorganising around AI at the top of the org chart, which is either empowering or clarifying, depending on where one sits in the org chart.
What happens next
The firm will continue scaling, the governance framework will continue governing, and the humans will continue being accountable for the outputs of systems they did not write and cannot fully inspect.
The lawyers have documented all of this carefully. That will prove useful.