Cybersecurity has become a household term in the recent times. As you may well know, the world was recently in turmoil after the “WannaCry” ransomware wreaked havoc in almost 150 countries.
We are increasingly living digital lives which is only going to get more pervasive in the future. Everything from your pet’s photos to your financial holdings will increasingly be stored online.
In such a world, the thought that anyone could access and steal even your entire identity, is a scary one. We cannot go back in time and live analog lives, technologically speaking, therefore, there needs to be cutting edge solutions for the security of our digital lives.
AI To The Rescue?
AI or Artificial Intelligence has become another household term in the recent times. Various tech companies and research groups are working on state of the art AI, which will be used in assisting us in various aspects of our lives.
We are already starting to see the immense usefulness of AI in various systems like in Tesla’s self driving cars that are increasingly getting smarter and preventing accidents, in our email systems to scan for and prevent phishing/junk emails, on our smartphones in the form of virtual assistants who by the way can now even control our homes.
The AI systems of the world are getting smarter by the second. In fact, it is pretty much how they work. The more data you feed into them the smarter they get. Some scientists and experts expect AI to be one day more intelligent than human beings that will be able to work independently from us. They are already beating us in board games and video games!
So, could a general purpose AI that is as or more intelligent than us, be useful in defending against any sort of hacking? Would it have the potential to end all hacking?
Computer Code Defending Against Computer Code
An AI is essentially computer code that is designed to perform certain tasks. However, an AI is adaptive in nature and it can quickly learn based on the data it is fed in real time.
So, in theory, an AI that is defending a system could learn from an attack as it is happening and could devise solutions to it and defend against it, all in real time!
Not only that, it would immediately adapt itself, ironing out the vulnerabilities in real time along with identifying the attacker and attacking them instead, all happening at once!!
The AI system that we have today are able to perform some of these tasks already but a truly intelligent AI would just be a beast unlike anything we can imagine today. But there is one problem. There is nothing stopping the attackers from using the same AI as well.
AI Vs AI
It gets interesting from here. We know that technology is a doubled edged sword and an AI is no different. It could help serve the righteous and it could help serve the evil.
So, in a potential scenario, the attackers could use their state of the art AI to attack a tech company’s system who is being defended by their state of the art AI.
Such a battle would be very interesting. While the defending AI would be learning in real time trying to avert the attack by any means possible, the attacking AI would be doing the same, learning in real time how its opponent is working and trying to devise ways around its defences.
That sounds pretty futuristic, doesn’t it? It would surely make up for an intense scene in a sci-fi movie. But getting back to the point, an AI is only useful for the purposes it is built and therefore that makes it more of a tool that is available to everybody and so, unfortunately, it will not be ending hacking. Now, if it ends humanity, it will have ended hacking indirectly, but I'm not rooting for that!!
All images used are taken from the public domain except for the first one.