The Sydney Morning Herald and The Age Removed an Opinion Article After Undisclosed Copilot Use

The Sydney Morning Herald and The Age removed an opinion article by Western Sydney University academic Cath Ellis after reporting found it had been prepared with Microsoft Copilot without the masthead being informed. Ellis said she used Copilot to structure her own notes, while the masthead said the article did not meet its editorial standards on AI use.

Risk classification

• Primary risk domain: 6 Socioeconomic & Environmental
• Primary risk subdomain: 6.3 Economic and cultural devaluation of human effort

The incident highlights the tension surrounding AI-generated content displacing human writing in journalism and academia, raising questions about authorship and the value of human effort.

Additional risk subdomains

• 5.1 Overreliance and unsafe use: The author relied on Microsoft Copilot to draft and structure her article, which violated the publisher's editorial standards and led to its retraction.

Causal factors

• Entity: Human
• Intent: Unintentional
• Timing: Post-deployment

The incident arose from a human decision to use a post-deployment AI system to draft a published article, leading to an unexpected retraction and public controversy.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The system used is Microsoft Copilot, a general-purpose AI chatbot and writing assistant. Under the EU AI Act, chatbots and systems that generate text are classified as Limited Risk (Level 3) because they require transparency obligations to ensure users are aware they are interacting with AI or that content is AI-generated.

AI system and alleged parties

• AI system: Copilot (Microsoft)
• AI purpose: Writing Assistant; Technical Text Generation
• Behaviour type: Assistant
• Alleged developer: Microsoft, Large language model developers, Chatbot developers
• Alleged deployer: Western Sydney University, Educational communities, Cath Ellis
• Alleged harmed parties: The Sydney Morning Herald, The Age, Journalism, Epistemic integrity, Educational communities

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Negligible. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Negligible, indirect Negligible
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Negligible, indirect Negligible
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Negligible, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Negligible, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

People affected

• Occurrences reported: 1
• People reportedly harmed: 1
• People reportedly exposed: 1

Potential causes

Management

• Uncommunicated AI Guidelines: Masthead guidelines against AI writing were never conveyed to the author.
• Permissive Institutional Policies: WSU management actively encourages and praises AI-assisted writing.

Technology

• Distinct AI Writing Patterns: Copilot generated recognizable patterns like rule-of-threes and jargon.
• AI Detection Unreliability: AI detection programs are useful but not foolproof for editors.

Data Inputs

• Large Volume of Input Text: Uploading 40,000 words of research to LLM led to unexpected phrasing.

Human Factors

• Overreliance on AI for Writing: Academic used Copilot to structure and write instead of drafting from scratch.
• Assumption of Acceptable Use: Author assumed using AI as a team member was appropriate for public pieces.

Process and Methods

• Undisclosed AI Use in Submission: Author did not inform the editors that AI was used to compile the article.
• Differing Institutional AI Norms: Discrepancy between university encouragement and media guidelines on AI.
• Lack of Pre-publication Check: Masthead lacked active checks to detect AI-generated text before publishing.

Regulatory Environment

• Lack of Uniform AI Standards: Absence of nationwide standards for AI use in education and publishing.

Information quality

• Classification confidence: High
• Reason for confidence: The report provides clear, firsthand accounts from the author, the university, and the publishers regarding how the AI was used, why the article was retracted, and the institutional policies involved. There is little ambiguity about the facts of the event.
• Alternative interpretations: The incident could be viewed purely as an academic integrity or publishing ethics issue rather than an AI safety risk, but the central role of the LLM in generating the text makes it a clear case of AI-related socio-technical friction.

An Australian academic used Microsoft Copilot to draft an opinion piece, leading to its retraction due to publisher policy violations. This incident highlights evolving norms around AI in professional writing but carries negligible national security implications.

• Overall national security impact: Negligible
• Scope: Single nation
• Primary target: No clear primary
• Alleged perpetrator: Individual

Threat characteristics

• Imminence: Long-term. This incident is a minor ethical dispute with no active national security crisis or threat vector.
• Autonomy: Human-controlled. The AI acted as a writing assistant, processing the author's own research notes under her direct supervision and control.
• Novelty: Established threat. The use of LLMs for drafting text is a highly common and established practice with well-known detection and policy challenges.

Impact by dimension

• Physical security: Negligible. No physical security threat, kinetic targeting, or critical infrastructure compromise occurred during this academic writing incident.
• Information security: Negligible. While involving AI-generated text in media, this was an authorized academic using AI to draft her own research, not a hostile disinformation campaign or intelligence compromise.
• Sovereignty: Negligible. No state authority, electoral processes, or core government functions were targeted or impacted by this incident.
• Economic security: Negligible. The incident represents a minor dispute over publishing ethics and has no impact on strategic technological advantage or economic security.
• Societal stability: Negligible. This incident relates to academic integrity and editorial standards, posing no threat to civil liberties, societal stability, or human rights.