A new attack method known as 'Ghostjacking' has emerged, where attackers manipulate logs to control AI agents. By embedding malicious instructions into logs or alerts that record blocked requests, these agents can be tricked into executing harmful commands. This poses a significant risk, particularly for organizations that rely on AI to automate processes or manage security alerts. The implications are serious, as compromised AI systems could lead to unauthorized actions and data breaches. Companies using AI technology need to be vigilant about how logs are managed and ensure they are protected against such manipulations.