The progression in Artificial Intelligence have incited intense debate worldwide, some experts are calling AI to counter malware attacks
In a short span of six weeks, the world was hit twice by major ransomware attacks — malicious software that seizes the data stored on your computer systems and would only release it to you upon receiving ransom money.
It goes without saying that the world is in dire need of better defense mechanism, and mercifully they have started to arise but in a rather slower and in patchwork fashion. The day we would feel completely secure, we may have artificial intelligence to thank.
Ransomware is not necessarily the trickiest or the most dangerous forms of malware that can hack into your computer, but it surely is infuriating and at many instances devastating. The majority types of infections that sneak into your systems don’t lock down your personal pictures or documents the way ransomware does and nor do they demand money.
Despite knowing the risks, there are many who aren’t very savvy enough to keep up with the security software updates. Both of the recent ransomware attacks thwacked users who were unsuccessful at installing the security update which Windows released a few months earlier.
The progression in Artificial Intelligence have incited intense debate worldwide. Science fiction is nearing reality and AI products are taking over households and workplace. Concerns about the potential takeover of AI over the job market is rising. Discussions are also live about the detrimental effects of an AI singularity, taking over the world and terminating the human race.
Though these debates are all valid, I strongly hold the opinion that AI should not only be concentrated to just home gadgets or on process optimization and automation. Instead, AI can be put to use in solving the greater problems the world is facing.
AI in tackling Ransomware
2015 saw around 707 million cybersecurity breaches while in 2016 the figure reached 554 million in just the first half.
Earlier, the identification of malicious programs was done by matching the program’s code against a database of identified malware. Though this method was only as good as the database and would ultimately get outdated, allowing new malware variants to easily sneak in.
Hence, security companies started to gauge a malware by its behavior. For instance, in the case of ransomware, the software may go for repetitive tries to lock files by encrypting them. But this could also flag regular computer behavior for example file compression.
The modern techniques now include considering a combination of behaviors. Such as, a program which is encrypting files without displaying a progress bar on the monitor could be signaled as a surreptitious activity. But the drawback here is that it slows down the process of harmful software identification, also locking up some of the files.
The growing number of alerts being generated by the traditional tools is only increasing the struggle for the security teams. AI, with its ability to self-learn and automate, can raise the effectiveness and cut cost, guarding us from terrorism or attacks of rather smaller scale.
Moving further, the existing AI-centric solutions in the industry are more pro-active. They have the ability to anticipate attacks beforehand by detecting patterns and glitches pertaining to malicious content.
Secureworks utilizes the predictive competencies of AI for advanced threat recognition globally. SiftScience, Cylance, and Deep Instinct are using it to prevent frauds and for endpoint security, like smartphones and laptops. These technologies hold the potential to radically magnify the possibilities and scale of security specialists and enable them to sense incoming threats before they actually materialize.
Working as a cyber security solutions architect, Alisa focuses on bug bounty and network security. Before joining us she held a cyber security researcher positions within a variety of cyber security start-ups. She also experience in different industry domains like finance, healthcare and consumer products.