Since the dawn of the Internet, back in the early 80s, #cybersecurity has been a major concern for businesses. For a long time, cybersecurity has been reduced to breaching of information (via phishing attacks) or information theft like credit card fraud and so on. Now, the scale and scope of it, have grown exponentially.
Two majors factors have helped influenced the magnitude to which cyber threats have evolved to:
Cloud Systems and the #IoT (Internet of Things)
Many companies are now adopting the #Cloud, as opposed to having a server room, to store their valuable information. With this new way of information management, comes new challenges. First, there is an element of trust that comes into play, with corporations having to rely on the Cloud System companies that will be handling most, if not all, of their data. Secondly, how secure is the Cloud computing environment? Again, organizations will have to be confident that the Cloud solution is a far more secure option than having a server room at the office.
Internet of Things: Attack Surface in Modern Digitalisation
With the Internet of Things, we now have almost, every single device connected to the Internet. From your smartphone to your smartwatch, or the office coffee machine and even your car (i.e. #Tesla) that is connected to the Internet, we get a very broad cyber threat environment, where most things can be made vectors to cyber attacks. With that, comes the challenge of knowing what's connected in a working environment.
For instance, Tesla's AI studies the way a person drives (average driving speed, braking patterns, etc...) and analyze the roads and routes taken by the driver. This information is then transferred back to Tesla's servers for analysis, to better the self-driving option of the car. So a Tesla car, parked in front of an office building could easily transfer its data to headquarters, while connected that office's wifi. Hackers could use the car as an opportunity to penetrate the office building's network.
Cyber-attacks are getting much more creative and complicated, that we can no longer rely on historical data to predict future attacks. You never know for sure where the attack will come from. Back in the days, we could anticipate that robbers who will want to rob a bank will either come through the front doors or attack the cash in transit.
The cyber threat environment is much different. The range of people or organizations getting involved is very broad; their motivations are almost random (some do it for fun, some for the money, others to prove they can do it, etc...); And the choice of attacks available is too extensive for companies and cybersecurity teams to predict them with accuracy.
The use of Artificial Intelligence and Machine Learning, as cybersecurity defensive measures, is becoming crucial. AI bots, through Machine Learning can be taught defensive algorithms that will easily detect and automatically respond to cyber-attacks. In many cases, Machine Learning responses can help get ahead of issues before they become serious threats. The scope and speed of threats are so fast that, humans do not really have a chance to respond.
The human body has its immune system to help fight viruses and recover for diseases. Likewise, AI through Machine Learning can get a true and precise understanding of the machine and will then be able to accurately find and provide a rapid response to arising threats. As AI learns about threats, it will be able to provide an autonomous response to attacks. But this will only be possible through building a human-AI trust relationship.
How can this relationship be built?
Start by building AI features that provide accurate recommendations on steps to take, each time there's a threat or breach in the system. The more issues are solved by those recommendations, the more trust is placed in the AI to handle situations.
There will come a time where humans will have no choices but to trust Artificial Intelligence, due to the rampant scale and sophistication of cyber-crimes. If we want to save on time, avoid considerable losses (in case of a security breach), mitigate cyberthreats risks, we need to invest in AI and Machine Learning to provide a safer cyber environment.
The question was asked: What if AI and Machine Learning are used to generate attacks? That's a question worth a whole post.