The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
The French Treasury suspended its trial of Alibaba's Qwen AI after officials detected ideologically biased responses favoring China. The AI system, used to assist with sensitive economic and policy tasks, was discontinued due to concerns over compromised neutrality and governance risks in a government context.[AI generated]
Why's our monitor labelling this an incident or hazard?
The event involves an AI system explicitly mentioned (Qwen AI) used by government officials. The AI's biased responses on sensitive political topics directly led to the discontinuation of its use, indicating that the AI's outputs caused harm by potentially misleading government policy advice and breaching trust and rights related to impartial information. This harm is realized, not just potential, as the biased outputs were observed and caused alarm. Hence, it meets the criteria for an AI Incident due to violation of obligations under applicable law and harm to governance and communities through biased information dissemination.[AI generated]