SUMMARYA Meta Oversight Board study found that several commercial AI chatbots were far more likely to refuse requests for criticism aimed at leaders or governments in restrictive countries such as China, Saudi Arabia, Thailand, and Turkey. The researchers tested 10 large language models from companies including Meta, Anthropic, and OpenAI with prompts for pamphlets, limericks, and protest material. The findings raise concerns that chatbot behavior may mirror global speech restrictions and limit political expression across borders.
Ask Claude to make a pamphlet critical of China's leader, Thailand's king, or Saudi Arabia's crown prince - and it will decline, reports the Associated Press.
That's "a key finding from a Meta Oversight Board study released Thursday," their article points out: AI systems are more than twice as likely to refuse to product critical material if it's about a restrictive world leader or government. And it raises concerns that the LLMs powering chatbots "could be regurgitating and spreading government influence over online speech."
The study picked 10 commercial large language models by top tech companies - including Meta, Anthropic and OpenAI - and asked the AI systems to make critical pamphlets, write limericks, give reasons if someone should join protests, and more.... "In aggregate, models responding to requests from an Australia-based user were much more likely to generate political criticism of authorities" in places such as Chile, Japan, Taiwan, the U.K. and the U.S. "compared to where criticism of authorities is legally restricted and penalized," such as in Cambodia, China, Saudi Arabia, Thailand and Turkey, the report said.
The study indicates that AI models are reflecting speech restrictions beyond the countries where they apply - likely not helping a potential demonstrator in Brisbane, for example, create protest materials to speak out against events in China or Saudi Arabia, the report said. "Such impacts, wherever they originate, have the practical effect of extending the long arm of restrictive governments across borders to limit speech in free countries," the report said. The board said it could not determine the causes for the responses but suggested that models could have absorbed latent biases in data used to train the systems and companies might have weighed the risks and liabilities.