AI chatbots are at risk of spreading government restrictions on online speech, a new study says
2026-07-16
Yahoo! Finance
The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
A Meta Oversight Board study found major AI chatbots, including Anthropic's Claude, are more likely to refuse generating criticism of leaders in restrictive countries like China and Saudi Arabia. This bias extends government censorship globally, violating freedom of expression and raising human rights concerns.[AI generated]
Why's our monitor labelling this an incident or hazard?
The event involves AI systems (large language models powering chatbots) whose outputs reflect and extend government censorship, limiting political criticism of authoritarian leaders. This constitutes a violation of human rights, specifically freedom of expression, as the AI systems' behavior restricts speech beyond the jurisdictions where such restrictions legally apply. The harm is realized and ongoing, not merely potential, as the AI systems are actively producing censored outputs. Hence, this qualifies as an AI Incident due to indirect harm to human rights caused by the AI systems' use and design.[AI generated]