Major publishers are tightening rules on AI-generated opinion pieces. New guidelines demand transparency from contributors. Recent scandals have forced outlets to delete columns and rethink editorial standards.
News organizations are moving quickly to address the growing risk of undisclosed AI-generated content on opinion pages, as recent incidents have raised concerns about editorial integrity and audience trust. Publishers are now implementing stricter guidelines and requiring contributors to disclose any use of artificial intelligence in their submissions.
In the past week, Deutsche Welle reported that two prominent German news outlets deleted opinion pieces after discovering the authors had used AI without disclosure. Berlin's Tagespiegel removed op-eds by its former publisher and editor-in-chief, Stephan-Andreas Casdorff, while Frankfurter Allgemeine Zeitung (FAZ) confirmed that a column by Thuringia state premier Mario Voigt was also created with AI assistance.
Other high-profile cases have surfaced internationally. The New York Times dropped freelance journalist Alex Preston in March after learning he used AI to write a book review. In April, the Mississippi Free Press revealed it had unknowingly published an AI-written opinion column attributed to a fake author. Earlier this year, Peter vanderMeersch, retired CEO of Mediahuis Ireland, was found to have published several AI-generated columns on Substack, some containing fabricated quotes.
Amid these developments, the editorial team at Altinget, a political news site with editions in Denmark, Norway, and Sweden, introduced five new guidelines for op-ed contributors. The editors said they have seen a noticeable decline in the quality and authenticity of debate, attributing this to the increased use of AI in submissions. They observed that AI-generated texts tend to be uniform in tone and structure, often lacking coherence and a personal voice.
Altinget's guidelines emphasize that the platform exists to facilitate debate between people, not machines. Contributors are encouraged to use AI only for brainstorming, research, or correcting grammar, but are expected to ensure that arguments and reasoning are their own. All factual information must be verified, and contributors must be transparent about any AI involvement. Posts that are mostly or entirely AI-generated will be rejected or sent back for revision.
These measures reflect a broader push for transparency and editorial standards in the face of rapidly advancing AI tools. As newsrooms adapt, some are looking to frameworks that prioritize human authorship and open disclosure. For example, USA Today's Jessica Davis has argued that newsrooms should measure AI performance with clear frameworks rather than relying solely on human oversight, a perspective detailed in her recent discussion on data-driven AI evaluation in newsrooms.
