Tech Giants Like Amazon, Google, Microsoft are Using AI to Fight HackersShare Tweet Share Pin it
AI (Artificial Intelligence) is a double-edged sword that can be dangerous in both ways. Yes, it can surely help to save lives by the early detection of potential diseases like cancer, if it goes into the evil hands, they can use it to launch potent attacks.
Last year, the Azure Security Team of Microsoft detected some unauthorized access in cloud computing. It was surely an attack by the hacker. With the immediate alert from Microsoft to the customer, the threat was prevented before the hacker can go too far in the system. This prevention was possible with the help of AI, of course.
After that incident, the tech giants like Google, Microsoft, and Amazon are leveraging more on Artificial Intelligence to tackle with such potential threats from the hackers. The CSO (Chief Security Officer) of each of these companies has spoken to the international news agencies and told that the ML (Machine Learning) and AI would play a crucial role in protecting the giant tech company’s multi-million dollar facilities by crushing large & unwanted data pool every day.
The Product Management Director of Google, Mark Risher while speaking on the large data sets that are needed to be processed for determining the unauthorized activities said that “The amount of data we need to look at to make sure whether this is you or an impostor keeps growing at a rate that is too large for humans to write rules one by one.”
With the help of Machine Learning, Microsoft has made improvements to bring down the false positive ratio to 0.001% by customizing as per the user’s history and online behavior. But with the smartness of technical security, hackers have become smarter. For example, an intelligent hacker can identify the algorithm on how the companies train their system and can then use this info to corrupt or evade it. Though the multinational companies are acknowledging that it is not possible or is extremely difficult to defend the system from every single intruder, they argue that with the help of this new technology combination, they can fight hackers and can do the favor for the right people.
Stephen Schmidt who is Amazon’s CISO (Chief Information Security Officer) has said that the industry will “get incrementally better at protecting systems and make it incrementally harder for attackers.” “We will see an improved ability to identify threats earlier in the attack cycle and thereby reduce the total amount of damage and more quickly restore systems to a desirable state.”, he added.
The Macie service of Amazon uses machine learning to find the location details and the number of access from the customer who is watching Netflix. With the help of AI and ML together, it is also possible to identify whether the user is browsing or searching through an Ethernet connection or a wireless network with a WiFi card. Even the broadband plan, connection speed, and currently active users on the web can be identified. The combination of AI and ML makes it possible to catch the unauthorized login to a system or network and then to deny the permission with a single yes or no or even without it.
By now, you may have determined that the security from Machine Learning may not work in all instances, especially when it has a small amount of data to train. The tech companies are worried about the danger of attacks by hackers on the systems with such a lack of threat protection. That’s why companies like Google, Amazon, and Microsoft are making their operations more reliable on Artificial Intelligence.
However, such threats are more on the table than in the real systems. But, that doesn’t mean that it cannot be changed. Because ‘Security is like a weapon race. Whichever side has more accurate, long-range armory along with the ability to defend own territory will have the upper hand in any kind of situation.’ That’s why the experts are advising such technology giants and also to all the other vulnerable companies to change their system formulas at random times and keep them secret too.