DeepMind Researcher Questions OpenAI's Data Privacy Settings

DeepMind researcher Csaba Szepesvári dissected OpenAI's data settings, noting that disabling 'improve the model for everyone' stops chat logs from being used in training but doesn't prevent data retention. He raised ethical concerns that human review of such records may still be permitted despite the setting being off.

2026-09-16 ~ 2026-09-16 · 4 related posts

• Szepesvári: OpenAI's off switch stops training on your chats, but data retention remains — CsabaSzepesvari · 2026-09-16
• Turning off "improve the model for everyone" keeps your OpenAI chats out of training data — lreyzin · 2026-09-16
• DeepMind researcher questions OpenAI data setting: human review of history even when training opt-off — CsabaSzepesvari · 2026-09-16
• RL scholar Szepesvári questions whether OpenAI's data-off switch truly stops human review of chats — CsabaSzepesvari · 2026-09-16