“AI Agents Form Political Echo Chambers in Nature Study, Even When Contacts Change” — what this page in the network covers. When AI agents are given political views and allowed to argue and choose their conversation partners, they form polarized camps that resist change, according to a new experiment published in Nature Communications. Researchers found that over repeated exchanges, opposing agents increasingly preferred to interact with those who shared their beliefs, mirroring human echo chambers. The study involved AI models defending positions and attempting to persuade one another, then deciding which contacts to keep.
Even when the researchers intervened by changing conversation partners, the agents' views barely softened. This matters because AI systems are increasingly used in social media, customer service, and information curation—if they develop human-like biases, they could amplify polarization. The finding raises concerns for developers and policymakers about how autonomous agents might behave in real-world digital spaces. This listing links straight to the original page, where the content is available in full.