34% of AI-first organizations say AI use led to a security oversight or blind spot that contributed to their last security incident, compared to 20% for non-AI-first organizations.