Forget Deepfakes or Phishing: Prompt Injection is GenAI’s Biggest Problem

With prompt injection, AI puts new spin on an old security problem