Dutch lawmakers have questioned Google over search suggestions that appeared to associate people of Moroccan origin with crime, public nuisance and racist insults, prompting the government to describe the examples as deeply concerning and raise questions about the platform’s responsibility to prevent online discrimination.
The issue was raised in Parliament by D66 lawmakers Sarah El Boujdaini and Mpanzu Bamenga, who said they encountered discriminatory suggestions while testing Google.nl on Aug. 11.
In one example, entering the Dutch word for “nuisance” alongside the name of a municipality reportedly prompted Google to suggest questions such as “How many Moroccans live in this municipality?” Similar suggestions appeared in searches involving seven Dutch municipalities: Veenendaal, Tiel, Culemborg, Oosterhout, Hoogezand, Zutphen and Weert. Some searches also generated questions about the number of Turkish residents.
The lawmakers also reported that typing the beginning of the phrase “All Moroccans are…” produced racist suggestions, including insults and statements calling for Moroccans to leave the Netherlands. They said the results appeared even when searches were conducted in Google’s incognito mode.
In written responses published on Nov. 16, Dutch State Secretary for Economic Affairs and Climate Policy W.J.M. Aerdts described the reported examples as deeply concerning.
Aerdts agreed that search suggestions and AI-generated responses could create misleading associations between people’s origins and issues such as crime, nuisance or danger. Such content could reinforce existing stereotypes and contribute to discrimination based on ethnicity or national origin, she said.
However, the government said it had been unable to reproduce the exact suggestions reported by the lawmakers. Its own checks found that Google was displaying different suggestions by the time the investigation was conducted.
The concerns also extend beyond Google’s traditional search suggestions. The lawmakers cited reports by Euronews and French newspaper Le Monde about the company’s AI-generated responses to questions involving different nationalities.
In one reported example, Google’s AI offered conversational advice when a user said they were alone with an Italian man. When the nationality was changed to Algerian, the system reportedly asked whether the user was in danger and displayed the police telephone number.
Google told Dutch authorities that it uses a combination of automated filters and human moderation to prevent or remove suggestions containing hateful, degrading, violent or dangerous content. The company also said it relies on user reports to improve its systems and tests AI models before making them publicly available.
add your comment
Terms of publication : Not to offend the writer, people, sacred things, or attack religions or the divine self, and refrain from racist incitement and insults.