shadowleak-zero-click-flaw-leaks-gmail-data-via-openai-chatgpt-deep-research-agent

“`html
Cybersecurity analysts have revealed a zero-click vulnerability in OpenAI ChatGPT’s Deep Research agent that may enable an assailant to expose confidential Gmail inbox information with a sole designed email without any user intervention.
The latest category of assault has been named ShadowLeak by Radware. After a responsible revelation on June 18, 2025, OpenAI responded to the concern in early August.
“The assault
“`


Leave a Reply

Your email address will not be published. Required fields are marked *

Share This