
The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
Researchers from Oxford and the University of Kentucky bypassed ChatGPT's safeguards to reveal the AI's embedded regional and racial biases, such as labeling Mississippi as the 'laziest' state. These findings highlight how AI systems can perpetuate harmful stereotypes, indirectly causing societal harm through biased outputs.[AI generated]
Why's our monitor labelling this an incident or hazard?
The event involves an AI system (ChatGPT) whose use reveals embedded biases that reflect and perpetuate harmful stereotypes, which can be considered violations of human rights and harm to communities. Although no specific direct harm incident is described, the article clearly outlines how these biases can influence users and societal perceptions, constituting a plausible risk of harm. The study's findings and the researchers' analysis indicate that the AI system's outputs have a pivotal role in these harms. Therefore, this qualifies as an AI Incident because the AI system's use has indirectly led to harms related to human rights violations and community harm through biased outputs.[AI generated]