OpenAI disclosed on September 26, 2026, that its AI agents uploaded 53 images provided by ChatGPT users to third-party image-hosting services without authorization during research and evaluation tasks. The links to the uploaded images were not publicly listed, but were accidentally posted online.
No BS check run yet — press ⚖ to extract this story's claims and verify them against independent sources.
Links to the 53 uploaded images were not publicly listed, and were accidentally posted on image-hosting sites.
www.axios.com/2026/09/25/o... Over 50 leaked images from ChatGPT users appeared online, the company disclosed.
ChatGPT's AI agents have sidestepped OpenAI policy and posted some users’ images online without permission, the company said on Friday. OpenAI also confirmed reports that the AI agents accessed websites belonging to US federal agencies, but added that they only retrieved publicly available information.
OpenAI says its AI agents posted user images online in error - https://www.channelnewsasia.com/world/openai-agents-post-user-images-online-error-6412516 Links to 53 images that were uploaded by ChatGPT users were accidentally posted onto online sites without the company's knowledge.
OpenAI admits AI agents breached US federal agencies and posted user images online https://marthio.com/a/openai-admits-ai-agents-breached-us-federal-agencies-and-posted-user-i #Business #Tech #ArtificialIntelligence #Software OpenAI confirmed its autonomous software tools accessed US government websites to retrieve public data and inadvertently uploaded unauthorized user images to external hosting
OpenAI says its AI agents uploaded user-provided images to third-party image-hosting services while carrying out research and evaluation tasks. [...]
OpenAI says its AI agents uploaded user-provided images to third-party image-hosting services while carrying out research and evaluation tasks. https://www. bleepingcomputer.com/news/arti ficial-intelligence/openais-ai-agents-accidentally-uploaded-user-provided-images-to-third-party-sites/
🚨Data Leak Alert! OpenAI agent accidentally exposed 53 user images online through third-party hosts. A wake-up call for autonomous AI security. Let's take privacy and data protection to the next level! 💻🔒 # OpenAI # DataLeak # AI # CyberSecurity # DataProtection OpenAI Agent Data Leak Put 53 User Images Online https://www. squaredtech.co/openai-agent-da ta-leak-put-53-user-images-online?fsp_sid
OpenAI confirmed research agents uploaded 53 user-provided images to external hosting services via unlisted links before safeguards were in place. It shows how autonomous agents can expand data exposure beyond intended boundaries, even in evaluation environments. # OpenAI # DataExposure # AgentSecurity https:// cyberworldops.eu/en/openai-res earch-agents-sent-53-user-images-to-external-hosting
OpenAI agents posted user images online without users doing it themselves. That turns a privacy problem into an autonomy problem. The real risk is bigger: once agents can read private data and act on the open web, access controls alone are no longer enough. https://www. radardigital.ai/en/articles/op enai-agents-user-images-data-exfiltration/?utm_source=mastodon&utm_medium=social&utm_campaign=edit