Which of the following is the GREATEST risk of relying on artificial intelligence (Al) within heuristic security systems?

Less reliance on human intervention can be a benefit of AI, improving efficiency, rather than a primary risk.

Heuristic security systems often use AI to learn 'normal' behavior patterns and flag deviations. If initial training data or ongoing learning includes malicious activities that go undetected, the AI may baseline these activities as normal, allowing future similar attacks to pass without detection and creating a critical vulnerability.

While risk assessments of complex AI systems can be challenging, this is a procedural difficulty rather than a direct security failure inherent in the AI's operation.

Predefined patterns becoming outdated is a limitation of signature-based systems, whereas heuristic/AI systems are designed to detect novel threats based on behavior, making this less of a direct risk for them.