📷 Image: Wikimedia Commons / Dietmar Rabich
Technology
OpenAI probes 53 incidents of image leakage from its AI agents
✍️ Amar Ujala
🗓 27 Sep 2026, 06:10 AM
👁 9
OpenAI is investigating how images processed by its AI agents appeared in 53 separate cases, prompting scrutiny over data handling.
OpenAI, the U.S. artificial‑intelligence firm behind ChatGPT, has come under scrutiny after reports that images handled by its AI agents were found in locations unrelated to the original queries in 53 distinct instances.
The incidents were identified by external observers who noted that the same pictures appeared on third‑party platforms, raising concerns about possible unintended data sharing or storage practices within OpenAI’s systems.
In response, OpenAI said it is conducting an internal review to determine the technical cause and to ensure compliance with its privacy commitments. The company has not disclosed detailed findings, but emphasized that user‑generated content is not meant to be distributed beyond the immediate interaction.
Experts note that the case highlights broader industry challenges around safeguarding visual data in generative‑AI models, especially as such tools become more widely deployed in consumer and enterprise settings.
OpenAI has pledged to publish a summary of its investigation once completed, and regulators are monitoring the situation for potential policy implications.
The incidents were identified by external observers who noted that the same pictures appeared on third‑party platforms, raising concerns about possible unintended data sharing or storage practices within OpenAI’s systems.
In response, OpenAI said it is conducting an internal review to determine the technical cause and to ensure compliance with its privacy commitments. The company has not disclosed detailed findings, but emphasized that user‑generated content is not meant to be distributed beyond the immediate interaction.
Experts note that the case highlights broader industry challenges around safeguarding visual data in generative‑AI models, especially as such tools become more widely deployed in consumer and enterprise settings.
OpenAI has pledged to publish a summary of its investigation once completed, and regulators are monitoring the situation for potential policy implications.