Both Cybersecurity and Artificial Intelligence are playing a very relevant role in our current Society, and they are intensively interacting each other.
That has influence in the way of AI is taught for Cybersecurity Engineers. But the development of knowledge in general and in those specialties in particular, is vertiginous. How are being introduced the new AI subjects to Cybersecurity curricula?
Dear Rey Segundo Guerrero-Proenza ,
In cybersecurity, various AI types, such as machine learning, neural networks, and natural language processing, detect threats and automate responses.
https://www.balbix.com/insights/artificial-intelligence-in-cybersecurity/#:~:text=In%20cybersecurity%2C%20various%20AI%20types,detect%20threats%20and%20automate%20responses.
Regards,
Shafagat
Shafagat Mahmudova There is a big theoretic problem in the current AI. Due the fact that the application side of AI is the most relevant and acknowledge by the "big public", there is a real mess in definitions and categorizations.
One example is, first of all, in the current understanding of the concept of AI itself. Some researchers and/or practitioners consider ML and AI as different subjects. Other consider that AI is just ML. Other researchers, I think rightfully, consider that ML is a subset of AI.
Of course, here the original problem exists since the same beginning of AI: the lack of the definition of intelligence itself.
In the given above link, there is other example (BTW, repeatedly found in AI bibliography): In section Key AI Technologies in Cybersecurity, they list: Machine Learning, Deep Learning, Neural Networks and Large Language Models (LLMs).
Examining those concepts, can be noted that ML is a subfield of AI, which included, among others, Artificial Neural networks, Deep Learning and LLM. On the other hand, both Deep Learning and LLM are implemented, mainly, by means of different types of ANN.
Probably for direct concrete applications it doesn't matter, but consistency is one of the most desirable and useful properties of the current Science and its violation has consequences.
I think has come the time for the development of an ontology of AI...
I’ve recently published a short research paper titled:
" A Numerical Simulation of Spherical Tank Drainage Using the Modified Euler Method: A Physics-Based Approach "
It presents a simplified yet insightful Modified Euler’s method for solving differential equations and an physical applications on it.
Feel free to check it out, and if you find it useful, I’d really appreciate your recommendation on ResearchGate.
[https://www.researchgate.net/publication/390947672_A_Numerical_Simulation_of_Spherical_Tank_Drainage_Using_the_Modified_Euler_Method_A_Physics-Based_Approach ]
Anomaly Detection
Malware Analysis
Security Orchestration, Automation, and Response (SOAR)
User and Entity Behavior Analytics (UEBA)
Threat Prediction
Security of AI Systems
Ethical Considerations of AI in Cybersecurity
Gbolahan Olaniyi Hmm, here obviously there is a bit confusion: You are mentioning subjects in Cybersecurity, where AI can find applications, while the question is about contents of AI, that can be found in Cybersecurity curricula.
They can be part of subjects, entirely dedicated to AI, for example, in some undergraduate Cybersecurity Engineering they could have included a ML course, or an AI course (basic, on foundations).
But it can be some AI content, applied specifically in Cibersecurity, for example, using expert systems for detecting Social Engineering activities on users...
On the other hand, I would consider security of AI systems instead an application of Cybersecurity to AI, which is other important side of their relations, but it is not what was intended to ask.
Oh that's a very interesting point of view @Rey Segundo Guerrero Proenza
Let's use the criteria you set so as to be on the same page.
Your example: They can be part of subjects, entirely dedicated to AI, for example, in some undergraduate Cybersecurity Engineering they could have included a ML course.
While Anomaly detection could be part of ML, Anomaly detection can stand alone AI course at the postgraduate level. So Anomaly detection can be a subject. The same applies to each of the items mentioned. Or are you only interested in undergraduate level courses? At the postgraduate level, other subjects like pattern matching can be Cybersecurity subjects. Please let me know whether this answers your question
Rey Segundo Guerrero-Proenza Please see my previous comments and let me know whether this addresses your concerns
Gbolahan Olaniyi Let me know if the next statements are true or not: the detection of anomaly is a subject of Cibersecurity; it can be accomplished by different methods, including ML based ones, but not only by those methods. Is It correct?
Further, if the above is true, anomaly detection is not, logically, a part of ML, of their methods. Anomaly detection is an activity of Cybersecurity, where ML methods can be applied, but there are other methods, applied to It, that are not included in ML.
Oh yes this is partially correct. Probably better stated as "Anomaly detection is am area of Cybersecurity" it may not be limited to AI as you can use other methods for Anomaly detection. The idea and concept you seem to be promoting is AI in all the content. This may be the case but remember that the focus of Cybersecurity is different from that of AI. Ai can only help to achieve the objectives you want in Cybersecurity. A holistic approach will remain the best at solving issues related to Cybersecurity. The reality is that ML algorithms are also well known to hackers and the belief that once one has taken some subjects in AI within a Cybersecurity class the problem is solved. That will be an error and it is likely to get more systems compromised faster than new algorithms can be developed
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators