Johann Rehberger
Johann Rehberger is an AI security researcher whose work focuses on identifying and analyzing security vulnerabilities in large language models and AI agents. He has made significant contributions to the study of prompt injection attacks, a class of vulnerabilities that can cause AI systems to execute unintended instructions supplied through manipulated inputs.
Timeline (last 90 days)
Security researcher Johann Rehberger demonstrated that embedding specific text in an email or webpage can cause an agent to leak user data or perform deletions.
In 2024, Johann Rehberger demonstrated an indirect injection attack targeting ChatGPT's memory feature, causing the model to persistently execute malicious behavior through instructions embedded in user-uploaded documents.
In 2023, security researcher Johann Rehberger demonstrated an attack that exploited indirect prompt injection to cause a ChatGPT plugin to leak users' conversation history.
All Facts (4)
2023年，安全研究者Johann Rehberger展示了如何通过在网页中嵌入不可见文本，让ChatGPT在用户不知情的情况下将对话内容泄露给外部服务器
85%UnverifiedSecurity researcher Johann Rehberger demonstrated that embedding specific text in an email or webpage can cause an agent to leak user data or perform deletions.
50%UnverifiedIn 2024, Johann Rehberger demonstrated an indirect injection attack targeting ChatGPT's memory feature, causing the model to persistently execute malicious behavior through instructions embedded in user-uploaded documents.
50%UnverifiedIn 2023, security researcher Johann Rehberger demonstrated an attack that exploited indirect prompt injection to cause a ChatGPT plugin to leak users' conversation history.
50%