According to the reading, "Top Strategic Technology Trends for 2023: AI Trust, Risk and Security management (Gartner report 2023)," 73% of organizations had deployed hundreds or thousands of AI models. Which of the following is not listed as a problem of using AI models in the article? A. Data can be compromised by benign mistakes or malicious data poisoning during AI model training. B. Personal data protection requirements prohibit the use of production/identifiable data in AI model training in most cases. So related risk must be identified and controlled. C. AI regulations are proliferating worldwide, mandating auditable practices that ensure trust, transparency and consumer protection. D. AI models produce predictable and intended outcomes only if the correct and complete training data is fed in the training stage