OpenAI has disabled a ChatGPT feature that allowed users to make individual conversations discoverable by search engines after a researcher compiled nearly 100,000 such chats that had been indexed by Google, according to an Aug. 5 report by 404 Media. The scraped dataset includes sensitive material ranging from draft non-disclosure agreements and confidential contracts to personal relationship discussions, underscoring the privacy risks created when users opt to share their exchanges publicly. Interactions with ChatGPT are private by default, but the optional “share” function generated a public webpage that Google’s crawler could index. OpenAI’s chief information security officer Dane Stuckey said the company has removed the feature and is working with search engines to purge the indexed content, calling security and privacy “paramount.” Google did not comment. Although OpenAI is attempting to erase the links, third parties already hold copies of the data, intensifying concerns over how easily user-generated AI content can be exposed once online.