Is AI the answer to outsmarting the cyber criminals and keeping your business safe?
Posted: 17 May 2017 | By Michael Garwood
Cyber-crime has become a major talking point in the past week – with medical institutions across 100 countries – including the the UK’s National Health Service (NHS), falling victim.
The ‘ransomware’ attack, caused chaos, resulting in computers and servers having to be shut down and appointments and operations cancelled, with hackers demanding fee’s to release files.
In an article in The Telegraph newspaper, Mikko Hypponen, chief research officer at the Helsinki-based cybersecurity company F-Secure, described the attack as “the biggest ransomware outbreak in history”.
With cyber-crimes on the rise, the importance of ensuring security levels are to a standard to which such threats are unlikely has (arguably) never been greater.
But with some of the world’s biggest businesses proving to be vulnerable to such threats (Yahoo, Vodafone to name a few), traditional security software firms are forever facing a desperate race to outsmart, out think and stay ahead of the criminals.
According to Bart Selman, an American professor of computer science at Cornell University, the answer lies with AI, by being able to detect and learn from different forms of threat itself.
“AI is learning to detect patterns which might reveal signs of an attack”, he told Access AI. “Networks are constantly bombarded with attempts, but the AI can learn and recognise potential dangers that maybe a traditional security software cannot.”
Sounding out the crooks
Selman, who has authored over 90 publications and a variety of conferences and journals in AI and computer science added that hackers are now becoming so sophisticated, they are even able to use computer generated voices to help pass security during phone calls.
Again, he noted the use of AI could help distinguish humans from machines, to help thwart new attacks.
“Cybercrime is becoming very sophisticated,” continued Selman. “AI can be used to detect if a voice is a real or a computer. You might think that would be easy, but it’s becoming more difficult. AI algorithms may be better at detecting a fake voice than a human.”