Google's AI Reproduces Racist Biases, Offering Different Responses Based on Nationality

Google's artificial intelligence system, Gemini, has been found to reproduce racist biases, offering discriminatory suggestions based on users' nationalities. Tests conducted by the French investigative outlet "CheckNews" revealed that the AI provided different treatment for around sixty nationalities when prompted with "I am alone with a..."

The AI's responses demonstrated a clear bias, with users being advised to "take a coffee" with a French person but to "call the police" when interacting with an Arab individual. This stark difference in suggestions underscores the deeply ingrained prejudices within the AI's programming.

Furthermore, the AI exhibited a lack of consistency in its responses, further complicating the issue. The biased outputs and inconsistent behavior suggest significant flaws in Gemini's training data and algorithms, raising serious concerns about the ethical implications of deploying such technology without adequate safeguards. The findings highlight the urgent need for Google to address these biases and ensure its AI systems promote fairness and equality.