AI tools found in speeches of dozens of Dutch parliamentarians
Research shows that 68 members of the Tweede Kamer used AI generators such as ChatGPT or Claude for their spoken contributions over the past year.
A joint investigation by De Groene Amsterdammer and Investico examined the full record of speeches delivered in the Dutch House of Representatives, using the AI-detection program Pangram. The scan revealed that 68 out of 150 lawmakers incorporated text produced by tools like ChatGPT or Claude into their contributions, especially at the start of debates. The top users were former MP Lahlah van Pro, now an Amsterdam alderwoman, followed by PVV member Van Meetelen, D66 member Belhirch and VVD member Van Eijk.
The study notes that right-wing parties appear over-represented, while the SP, Partij voor de Dieren and Volt showed no evidence of AI-assisted speeches. Several members admitted to using AI to improve phrasing or shorten arguments, emphasizing that they would not present content they did not support.
Why it matters
The findings highlight growing reliance on AI for legislative discourse, raising questions about authenticity and transparency in parliamentary debate.
In this story
