Which of the following strategies BEST mitigates the risks associated with exploiting the capabilities of generative AI for cyberattacks?

Awareness campaigns educate users about risks but do not technically prevent attackers from exploiting generative AI capabilities against an organization's systems.

Hallucination prevention controls address AI reliability and accuracy issues but are unrelated to blocking cyberattacks that exploit AI capabilities.

Reducing AI use is an impractical and overly broad response that does not address the specific technical vulnerabilities exploited in AI-enabled cyberattacks.

Generative AI systems can be weaponized through techniques like prompt injection, adversarial inputs, and data poisoning, all of which rely on malicious data reaching the model. Implementing robust data validation at input and output boundaries detects, filters, and blocks malicious content before it can be processed or cause harm. This is a direct technical countermeasure that addresses the attack vector rather than relying on awareness or usage restrictions.