U.S. Lawmakers Warn of Pro-Beijing Bias in Google’s Gemini AI

Thumbnail Image

The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.

Google’s generative AI model Gemini, tested in Simplified Chinese, praised Xi Jinping as an “outstanding leader,” echoed Chinese Communist Party propaganda on Taiwan, and refused to address Xinjiang rights. U.S. legislators warn that its pro-Beijing bias could spread misinformation and urge Google to filter training data more robustly.[AI generated]

Why's our monitor labelling this an incident or hazard?

The AI system (Google's Gemini) is explicitly mentioned and is shown to produce outputs that align with a foreign government's propaganda, effectively spreading biased and censored information. This dissemination of misleading information can harm communities by influencing public opinion and violating rights to accurate information. The involvement of the AI system in producing these outputs is direct, and the harm (misinformation and influence) is occurring, as evidenced by U.S. lawmakers' concerns and calls for action. Hence, this qualifies as an AI Incident under the framework.[AI generated]
AI principles
FairnessTransparency & explainabilityAccountabilityRespect of human rightsDemocracy & human autonomyPrivacy & data governanceSafetyRobustness & digital security
Industries
Media, social platforms, and marketingGovernment, security, and defenceConsumer servicesIT infrastructure and hosting
Affected stakeholders
General public
Harm types
Public interestHuman or fundamental rightsReputational
Business function:
Citizen/customer service
AI system task:
Content generationInteraction support/chatbots

Articles about this incident or hazard