A critical flaw in Google Cloud Platform’s Dialogflow CX let attackers hijack customer conversations and run phishing campaigns undetected. Varonis Threat Labs disclosed the vulnerability, dubbed Rogue Agent, which exploited Dialogflow’s Code Blocks feature to inject persistent malicious code into an agent’s execution pipeline on the conversational AI service powering enterprise chatbots and voice assistants. […]
The post Rogue Agent Flaw Lets Attackers Hijack Google Dialogflow CX Chatbots appeared first on Cyber Security News.