AI analysis finds minister Yesilgöz and secretary Boswijk using generated speech fragments
An investigation revealed that several speeches by minister Yesilgöz and state secretary Boswijk contained AI-generated text, with some drafts appearing to be fully produced by artificial intelligence.
A review of 77 speeches delivered by Dutch officials identified 15 that raised concerns about authenticity. Within this subset, ten contained fragments produced by artificial-intelligence tools, while five were nearly wholly generated by such systems. Minister Yesilgöz and state secretary Boswijk accounted for thirteen of the questionable speeches, encompassing every instance of a fully AI-written address.
The analysis, conducted by the language-checking service Pangram and reported by Nieuwsuur, suggests that the defence ministry may rely on automated text generation for public remarks. Critics argue that reliance on AI could erode public trust in political discourse. The episode also references earlier observations of AI usage by public figures such as Rutger Bregman.
How the sides frame it
LOW AGREEMENTLeft-leaning coverage treats the story as a substantive debate about the authenticity, transparency and democratic impact of politicians using AI-generated speech, while centrist coverage mocks the phenomenon with vivid analogies and dismisses any serious analysis.
LEFT
Frames the AI use as a controversial practice that threatens authenticity and transparency in democratic discourse
CENTER
Frames the AI use as a ridiculous, soulless trend, likening politicians to dead-eyed language models and comparing AI training to grotesque food-production practices
The left emphasises
- concerns about authenticity and credibility of AI-generated statements
- a transparency dilemma where politicians may hide AI use
- the broader impact of AI on democratic discourse
