GovernanceJuly 21, 20264 min read

Atlassian Says a Human Must Be Publicly Accountable for AI Writing. Here's How You Prove It.

Atlassian's guidelines for writing with AI require a human byline on everything they publish, and say that human is publicly accountable for the content. Accountability is a promise until you can show who reviewed what. DraftView turns that promise into a verifiable record.

Atlassian Drew the Right Line

In its guidelines for writing with AI, Atlassian sets a clear rule: everything it publishes needs a human byline, and that human is publicly accountable for the content, including anything AI wrote or influenced. Facts, data, and quotes an AI produces must be verified by that accountable person before they ship.

It's the right line to draw. AI can draft the words, but a person has to own them. That principle scales from a single blog post to an entire documentation set generated by AI and signed off by a team.

There's just one gap. Accountability that lives only in a byline is a promise. When someone asks you to show that the accountable human actually reviewed the content — a customer's security team, an auditor, a regulator — a name on a page doesn't answer the question.

A Byline Names the Person. It Doesn't Prove the Review.

Publishing an author's name says who is responsible. It says nothing about what they did. Did they read the whole thing? Did they catch the sentence the model got confidently wrong? Or did they skim it and approve?

This matters more for AI-written content than for human-written content, because the failure mode is different. A human writer who is unsure tends to hedge or flag it. An AI writes every sentence with the same even confidence, including the ones it got wrong. The review is the only place that error gets caught — so a rubber-stamp review catches nothing while still producing a byline that looks exactly like a careful one.

If your evidence of accountability is a name, you can't tell the careful reviews from the rubber stamps. Neither can anyone you're trying to convince.

What Verifiable Accountability Looks Like

Real accountability answers specific questions about each sign-off:

  • Who took responsibility, and in what role.
  • What they actually did: the comments they left, the edits they made, the parts they changed.
  • Against which version: the exact commit the sign-off applied to, so a later push can't quietly invalidate it.
  • With what build status: whether CI was green when they signed off.

The distance between "Jordan is the byline on this page" and "Jordan, an approver, signed off on commit a3f9c2 after leaving four comments and three edits, with CI green" is the distance between a claim and a record.

The Record DraftView Keeps

DraftView is built around the review of documentation PRs — including the ones an AI opened. When an approver completes a review, DraftView writes a sign-off record: who signed off, the role they held at that moment, the head commit they signed off on, and the CI verdict at the time. The record is append-only, so it stays as durable history instead of a field that gets overwritten on the next change.

Alongside the sign-off, DraftView measures whether the review was substantive. It looks at what the reviewer did, not how long the tab was open. An edit or a comment counts as real engagement; time on the page alone does not. A sign-off with zero engagement is the signature of a rubber stamp — exactly the thing a byline hides and this metric surfaces.

And because a sign-off is only as good as the thing it signed off on, DraftView ties every sign-off to a specific commit. If the author pushes more commits after approval, any CI verdict tied to the old commit clears, and the record shows the approval no longer matches the head. The accountable human is on record for what they actually reviewed — not for whatever landed afterward.

You can export all of it as an audit record for a given PR: the reviewers, what each one did, the sign-offs, and the version each applied to. That's the document you hand over when someone asks the accountable human to show their work.

From Byline to Evidence

Atlassian's rule — a publicly accountable human behind every published piece — is the standard AI-generated documentation should be held to. DraftView makes that standard verifiable. The review happens in an interface a person can actually read, the sign-off is an explicit human decision, and the record shows that a named person made that decision against a specific version.

A byline says someone is accountable. DraftView proves it.

Make "publicly accountable" mean something you can show.

DraftView gives reviewers a clean interface to read and sign off on docs PRs, and keeps the record behind every approval: who signed off, on which commit, with what CI verdict. Export it when someone asks the accountable human to show their work.

Start your free trial

14-day free trial, no credit card required

Resources