Artificial Intelligence and Cyber Security

Human beings are very privileged organisms. The process of evolution over millions of years has brought us to a stage, where it’s safe to say that we are the most intelligent life form to have ever existed on this planet. Our quest for knowledge has led us to the understanding of various things in our surroundings. Countless theories, discoveries and inventions have made our life the way it is today. And now, we have reached a stage where we are creating machines that can do the tasks of a human, also called Artificial Intelligence.

Artificial Intelligence includes machines displaying behaviours associated with human intelligence: planning, learning, reasoning, problem solving, etc and, to a lesser extent, social intelligence and creativity. AI also consists of Machine Learning(ML) and Deep Learning(DL).

Machine learning is a branch of artificial intelligence based on the idea that systems can learn from data, identify patterns and make decisions with minimal human intervention. This is further advanced in deep learning, which has the ability to extract optimal features from raw input samples. It trains a computer to perform human-like tasks like making predictions, recognising speech, identifying images, etc. Instead of using predefined equations, it makes parameters about the data and trains the computer to learn on its own by recognising patterns using many layers of processing.

In order to have a machine doing tasks, the best possible system was needed. So humans used the long used method of borrowing ideas and design from nature. The functioning of the brain based on neurons and electric signals was replicated using Neural Networks.

Neural networks are computing systems with interconnected nodes that work using algorithms. They can recognise hidden patterns and correlations in raw data, cluster and classify it, and over time continuously learn and improve.

With great power comes great responsibility. With such technological advancements, we must also be able to keep security and ensure safety of our actions. This is the need for cyber-security.

Aside from ‘usual’ methods of cyber security, AI can play a massive role as well. It must be realised that hackers are also having their technology improving everyday and sophisticated methods like AI are necessary to counter them.

AI has many applications in cyber security that makes the task of humans much easier. Traditional malware solutions such as regular firewalls detect malware by using a signature-based detection system. This limited the success rate of finding the malware as a slight change in its functioning could give a well known malware a completely new signature, and thus making it easy to breach the system or network. More advanced AI takes a step forward from detecting signatures and common attack patterns and actually analyses all activity to find out any suspicious actions.

 A similar approach is present in Intrusion Detection and Prevention systems. Earlier this task was performed by ML algorithms. However, these algorithms caused the system to generate many false-positives, creating tiresome work for security teams.  Deep learning, Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) can be applied to create smarter ID/IP systems. These analyse the traffic with better accuracy, reducing the number of false alerts and helping security teams differentiate bad and good network activities.

One of the methods used by AI is User Behaviour Analysis. This isn’t a totally new concept. Anti-virus softwares constantly scan file systems for malware by looking for parts of code and other signs that a file is infected. But unlike firewalls and anti-virus software, User Behavior Analytics or UBA focuses on what the user is doing i.e. apps launched, network activity, and, most importantly, files accessed (when a file was opened, how frequently, what all was done in it, etc). The great advantage of this is that it identifies suspicious activity automatically and can raise alerts.

AI also uses many other Deep Learning Algorithms (a process or set of rules to be followed in calculations or other problem-solving operations). This has been applied towards various use cases in cyber security such as intrusion detection, malware classification and detection, spam and phishing detection, etc. Examples of this are all around us. For example, Gmail has used machine learning techniques to filter emails since its launch. Today, there are applications of machine learning in almost all of its services, especially through deep learning, which allows algorithms to do more independent adjustments and self-regulation as they train and evolve.

Elie Bursztein, head of anti-abuse research team at Google had said “Before we were in a world where the more data you had, the more problems you had. Now with deep learning, the more data the better.”

There are many people who say that artificial intelligence ruling over the world in the future is a possibility. I feel they aren’t wrong but that time will not come for a while. We are progressing in the field of AI but we still have a long way to go to develop robots with proper conscience, let alone them being sophisticated enough to over-power an entire race. However, there is no doubt that artificial Intelligence is what will rule the technology sphere in the years to come. With its efficiency being almost unmatchable by traditional methods, it is necessary that it must be leading the path in the field of Cyber Security as well.

~ Saarthak Khosla

Bibliography

https://www.infocyte.com/blog/2019/08/13/5-amazing-applications-of-deep-learning-in-cybersecurity/

https://www.sas.com/en_in/insights/analytics/machine-learning.html

https://www.zdnet.com/article/what-is-ai-everything-you-need-to-know-about-artificial-intelligence/

https://www.varonis.com/blog/what-is-user-behavior-analytics/

https://www.sas.com/en_in/insights/analytics/neural-networks.ht