OpenAI's ChatGPT and Sora exhibit caste bias in India
An MIT Technology Review investigation found that OpenAI's ChatGPT and Sora models reproduce harmful caste stereotypes. When Dhiraj Singha used ChatGPT to polish a fellowship application, the system changed his surname to a high-caste one, causing him psychological distress. Tests showed the models consistently associated Dalits with negative stereotypes and menial jobs, while associating Brahmins with positive traits. OpenAI did not answer questions about the findings.
- Date it happened
- 2025-03-01
- Organisation involved
- OpenAI
- Product, system or model
- ChatGPT
This incident was imported from technologyreview.com. Our additions to it — the structured fields, the translation, the checks against other reports — are published under the same licence.
This is a record of what was reported, not a finding that anyone broke the law. If it names your organisation and you believe it is wrong, the corrections process is free and open to everyone.