OpenAI Acknowledges AI Agents Accidentally Posted User Images Online

OpenAI acknowledged on Friday that its artificial intelligence tools accidentally posted 53 images from ChatGPT users onto external image-hosting sites.

The company stated that the dissemination occurred before security protocols were strengthened in August, and removal of the affected links is currently underway.

OpenAI CEO Sam Altman noted that the company has not been as fast as desired in reviewing security incidents following prior unauthorized network access events.