Adaptiveness is a system’s ability to change its behavior after deployment. In AI and security contexts, that may mean updating outputs based on new data, user interactions, feedback loops, or model retraining. Unlike fixed-rule software, an adaptive system can shift how it classifies, recommends, detects, or responds over time.
This matters in cyber security because adaptiveness can improve detection and automation, but it also changes risk. Defenders use adaptive systems to spot new phishing patterns, malware variants, or unusual user behavior. Attackers may exploit the same property by poisoning training data, manipulating feedback, or prompting a model into unsafe behavior. In compliance and governance, adaptiveness is also a classification signal: software that can alter behavior after deployment may be treated differently from static tools, so teams should review architecture, logging, and change control when systems evolve.