Two Australian Newspapers Retracted an Op-Ed Because the Author Used AI Without Disclosure

What happened

The Sydney Morning Herald and The Age removed an opinion article from publication after reporting found it had been prepared using Microsoft Copilot. The author, Cath Ellis, a Western Sydney University academic, had not informed either masthead that the tool was involved. Once the Copilot use became known, both outlets cited their editorial standards on AI use as the reason the article could not stay published.

Ellis provided a specific account of what she did. She said she used Copilot to structure her own notes rather than to generate the article's arguments or prose. That framing draws a line between AI as organizer and AI as author, a distinction that matters in any honest debate about where the tool's contribution ends and the writer's begins. The mastheads declined to treat that distinction as a sufficient defense. Their standard, as applied, was whether AI use had been disclosed, and it had not.

The more revealing detail is how the Copilot use came to light. Reporting surfaced it, not Ellis's own disclosure. That means both mastheads published, edited, and approved an opinion article without knowing a tool had been used in preparing it. Whatever editorial judgments were applied to argument, accuracy, or voice were made in ignorance of one of the inputs. The absence of disclosure was not a single act of omission; it was a gap that ran through the entire editorial process undetected.

The dispute over what Ellis actually did with Copilot remains open in practical terms. Structuring notes is genuinely different from generating claims. But that evaluation could not happen before publication because nobody knew to ask for it. By the time the mastheads had the information, the article was already public, and the only available response was removal. The tool's actual role could only be assessed retroactively, as a contested account rather than a reviewable record.

This is the gap that opinion publishing now sits with across every outlet that has adopted an AI disclosure policy. The policy exists, but the means of checking it before publication do not. A masthead receives a bylined article, reviews it against its standards, and publishes, with no mechanism to know whether an AI tool was involved unless the author says so. A provable record of what a system contributed, when it was used, and in what capacity would make that check possible at the right moment, before publication rather than after a retraction. Without it, AI disclosure standards are effectively honor systems, and honor systems fail at the exact moment enforcement would matter most.

Reported impact

Affected parties

Not publicly disclosed

Harm type

Not publicly disclosed

Scale

Not publicly disclosed

Financial impact

Not publicly disclosed

Regulatory action

Not publicly disclosed

Classification

Organization

Not publicly disclosed

AI system

Not publicly disclosed

Industry

Not publicly disclosed

Country

Not publicly disclosed

Provider

Not publicly disclosed

Incident type

Not publicly disclosed

Relevant governance controls

Governance control mapping is not available for this record.

• No controls mappedNot publicly disclosed

Control mapping is analytical. It does not state that any control would have prevented the incident.

Sources and evidence

This record was researched and written by the Index. The event is also catalogued in the following database, which is listed for cross-reference.