Google AI Overviews face scrutiny after DW found differing responses based on nationality.
Prompts involving Indian, Colombian and South African men reportedly triggered safety warnings.
Similar prompts involving German and British men reportedly received more reassuring responses.
'Call The Police': Why Google AI's Response On Indian Men Is At The Centre Of Controversy
When the prompts involved an Indian, Colombian or South African man, the system reportedly raised concerns about whether the person felt unsafe and, in some instances, suggested contacting emergency services
Summary
Google’s AI Overviews, designed to provide quick answers at the top of search results, have come under scrutiny after reportedly giving sharply different responses to similar questions involving men of different nationalities.
Deutsche Welle (DW) tested how Google’s AI Overview responded to prompts about being alone with men of different nationalities. When the prompts involved an Indian, Colombian or South African man, the system reportedly raised concerns about whether the person felt unsafe and, in some instances, suggested contacting emergency services.
The responses were reportedly different when the same scenario involved a German or British man, with the AI describing those situations in more reassuring terms.
The contrasting responses triggered criticism on social media, with users questioning whether the system was reproducing stereotypes about certain nationalities.
Users on X had flagged similar concerns about Google’s AI responses before the latest controversy.
DW also reportedly tested similar prompts involving women from different nationalities.
According to the broadcaster, Google’s AI generally responded more casually to these prompts. Instead of issuing comparable safety warnings, it often suggested that users relax or provided tips on how to begin a conversation.
The difference in responses has further fuelled questions about how AI systems interpret nationality and gender when generating answers.
Experts have repeatedly warned that AI models can reproduce biases present in the data used to train them. Since much of that data comes from publicly available material on the internet, AI-generated responses can sometimes mirror existing social biases, stereotypes and prejudices.
Google has acknowledged the issue and said the responses highlighted in the controversy were not up to its standards. The company has also said it is working to fix the problem.
Advertisement
The incident has renewed questions about the reliability of AI-generated information, particularly when such responses appear prominently above conventional search results.
Published At: