Artificial intelligence (AI) is finding increasingly broad applications. It is present in many areas of everyday life – from online shopping to healthcare. Unfortunately, cybercriminals also use it, and thanks to these mechanisms they strengthen the effectiveness of their attacks. To fight back effectively, cybersecurity teams should therefore focus on understanding these techniques and use that knowledge to design their action strategy.
As early as 2018, the Electronic Frontier Foundation warned about the possibility of malicious use of artificial intelligence. And indeed, these mechanisms are increasingly becoming a major threat to digital security, because they allow cybercriminals to need less time to break into a network. Usually this operation took even several months, while now it takes a few days. Because of such high effectiveness, the number of attacks based on artificial intelligence and machine learning is growing. Shortening the time needed to carry out malicious activities also translates into lowering the associated costs. In this context, administrators' situation is made more difficult by cybercriminals' ability to automate the entire process of mapping networks, discovering potential targets, finding security gaps in them, and even carrying out non-standard attacks.
– Artificial intelligence in the hands of attackers poses an increasingly serious threat to the security of networks and data. The situation is made worse by their ability to use so-called intelligent swarms and to create automated, script-based threats. All of this is increasing the speed and scale of cyberattacks at breakneck pace – assesses Derek Manky of FortiGuard Labs at Fortinet.
The need to use an integrated security architecture
Integrated security architecture
In many enterprises, the protection architecture is not adapted to defend against AI-based attacks. The main problem is the use of even several dozen point network security products in a single environment, often from different vendors. This makes it difficult to obtain a full picture of the level of protection, because data from many applications must be integrated manually. In such a situation, an effective and coordinated response to a cyberattack covering the entire network is impossible.
Moreover, cybersecurity teams often cannot keep up with detecting threats, because criminals are becoming faster and minimize the time needed to carry out attacks. This in turn creates the so-called Breach Detection Gap, defined as the period from the moment of breaking into a network until the incident is detected. Data presented in the Ponemon Cost of a Data Breach Report 2020 indicate that this time can average as much as 280 days. According to that report, the average cost of a data breach worldwide is $3.86 million. The scale of this phenomenon shows how important it is to integrate existing security measures.
