Expert systems, or rule-based artificial intelligence, are systems with predefined behavior. Unlike AI based on machine learning, expert systems are not blank slates or tabula rasa as we say in Latin.
In expert systems, the environment is defined in advance and set to follow the rules established in that environment. If a certain action occurs, the system is meant to respond based on predefined parameters. An example of an expert system is chess robots or bots found in video games.
Yes, expert systems can be used in cyberattacks through the following methods:
Expert systems can be used to scan networks for vulnerabilities by following a sequence of predefined rules. These rule sequences can be used to find weaknesses in systems and network configurations.
Expert systems can be used to create a set of rules to analyze and gather information about target individuals. In this way, expert systems can help create more advanced phishing messages. When combined with Generative Artificial Intelligence (GAI), it can become very convincing.
Once vulnerabilities have been identified (for example, through machine learning), cybercriminals can instruct expert systems to follow a set of rules to exploit these vulnerabilities and carry out attacks.
Yes, expert systems can definitely be used in cyber defense by applying the following methods:
Expert systems can monitor network traffic and use predefined rules to detect unusual activities that may be a sign of a cyberattack. This method is useful for defending against known threats that are already recognized and registered.
Expert systems can identify potential security holes by analyzing system configurations and comparing them with known vulnerabilities. Antivirus or firewalls with a built-in expert system can quickly detect known threats, as well as typical signs of threats that resemble known threats. No firewall or antivirus will necessarily use only machine learning or expert systems, but rather a combination of both.
Expert systems can automatically take measures to isolate infected systems and block suspicious traffic when it detects a known threat or signs of a typical threat.
Several of Sicra’s partners have implemented artificial intelligence, including expert systems, in their solutions. If you want to learn more about solutions, we recommend the main article on artificial intelligence.
When it comes to becoming compliant with NIS2 and AI, we can assist through our CISO-for-Hire service.
Read about our "CISO-for-hire" service here >
Related words: Artificial intelligence, Compliance, Machine learning (AI), Neural networks and deep learning (AI), General artificial intelligence (GAI), GDPR, NIS2, NSM, DORA, ISO/IEC27001, CISO-for-hire.