Is Rob Jetten real or AI?

An in-depth investigation by de Volkskrant reveals that Prime Minister Rob Jetten and his party D66 have had artificial intelligence generate hundreds of social media posts since June 4, 2025. Of the 232 unique AI-generated posts, spread across 434 channels, as much as 30% of all posts by Jetten and D66 came from tools such as ChatGPT or Claude. De Volkskrant used the reliable AI detector Pangram to analyze the texts and concludes that using AI without transparency undermines the Prime Minister's authenticity.

Hundreds of posts, one question: who is the real Jetten?

De Volkskrant's investigation focused on 1,812 unique posts by Jetten and D66 since 2021. It found that AI was used not only for social media, but also for a page and a half of Jetten's book Hoe het wél kan (How It Can Be Done) and for texts in three campaign videos. Experts such as Claes de Vreese (University of Amsterdam) warn that language models create a "desirable emotion" rather than necessarily what the politician actually feels. "People will wonder: who is the real Jetten?" Suzan Verberne (Leiden University) calls the AI texts "bland uniformity" — generic, lacking spontaneity. Marc van Oostendorp (Radboud University) speaks of deception: "This is the opposite of transparency."

D66: "AI is a tool, not a deceiver"

D66 defends its use of AI as a tool for efficiency. The party uses AI for first drafts, spell-checking, simplifying difficult words, and shortening long texts. "Every post begins and ends with review and edits from staff," the party says. Still, D66 acknowledges that not all AI texts are edited and that errors slip in — such as illogical sentences about the housing chain or incorrect metaphors about the energy grid. In its 2024 European election manifesto, the party advocates for AI labels but does not apply them to its own posts because they have been "adjusted or approved" by a human.

AI texts recognizable by style and errors

De Volkskrant discovered that AI texts from Jetten and D66 stand out due to specific rhetorical patterns. The "rule of three" (three consecutive arguments) appears in 80% of the AI texts, while contrasts occur more than twice as often as in human-written texts. The posts also avoid em dashes to make AI use less noticeable. Still, illogical elements regularly appeared, such as a proposal to let prisoners work outside prison without being released from their cells. These errors and patterns make the texts recognizable as AI-generated.

Transparency paradox: labeling leads to distrust

The EU AI Act, in force since August 2, 2026, requires transparency about AI-generated content. Yet research shows that while 70% of consumers want transparency, that same transparency lowers purchase intent due to concerns about privacy and loss of control. IO+ calls this the "transparency paradox": labeling leads to rejection, even when the text has been checked by humans. The EU AI Act makes an exception for content that has been reviewed by a human with editorial responsibility — as D66 claims is the case.

This article, too, was written with AI, based on instructions from the editorial team and subsequently reviewed by the editorial team. AI as a writing tool is here to stay. But that makes the "what is real" debate all the more difficult.

Political communication in the AI era

The use of AI in political communication is not unique to Jetten and D66. During the 2026 municipal elections, 90% of AI-generated campaign content carried no label, despite the legal requirement. The PVV was the largest producer of AI content, followed by other parties. The Netherlands Institute for Human Rights warns that AI facilitates disinformation and manipulation, such as incorrect chatbot answers or AI-generated images without disclosure. The question is whether media and politicians can combine transparency with maintaining trust. Experts advocate for a cultural shift in which AI is seen as a legitimate tool, provided that ultimate responsibility remains with humans.