Free Trial

Machine Learning: a New Frontier in Mobile Phishing Protection

By Colm 21st November 2018

Targeting the ‘Mobile Workforce’

Mobile devices today are particularly susceptible to cyber attacks as more and more companies evolve into a ‘mobile workforce’. Employees that use mobile devices for both personal and professional business receive communications through an increasing number of channels. This means that instead of just email, hackers might target phishing attacks through SMS, WhatsApp and Facebook Messenger. Social media sites like Twitter, Facebook and Instagram also pose threats.

As well as this, employees are constantly on the move. They connect to countless Wi-Fi and mobile networks exposing themselves to threats. Cyber criminals are aware of all of these factors. As a result, they are constantly trying to exploit mobile users with phishing sites, malware downloads and social engineering attacks. As well as this, the current cyber security landscape is constantly changing and evolving. Therefore,  it can be difficult to anticipate and protect from potential threats.

Zero-Day Phishing Attacks

One of the most concerning examples of this constant attempt at attack is the development of ‘zero-day’ phishing attacks. Due to the real-time, constantly connected nature of mobile, phishing attacks continuously develop and evolve. Today, hackers create, deploy and dissolve phishing attacks all in as short as a single day. By publishing phishing sites online for such a short period of time before moving to an entirely new hosting server, hackers can easily evade detection while users are left with hardly enough time to identify, let alone prevent, the attack from occurring.

Businesses and individual users must be constantly aware of new emerging risks and attacks. However with over 46,000 new phishing sites created per day, with the majority of these online and active for only 4 to 8 hours, this is simply impossible. Where this human knowledge fails, users usually rely on anti-phishing and cyber security solutions. However, in the case of these ‘zero-day’ campaigns, often the threat has done its damage and moved on before it can be detected. The security software is too late to the crime scene . It is in these initial few hours that users are most at risk and mobile devices are most vulnerable.

Mobile Machine Learning

A major breakthrough in this attempt to anticipate and prevent cyber-attacks is the development of mobile machine learning solutions. Cutting-edge Artificial Intelligence (AI) technologies have made it possible to detect and react to threats as they are created. Therefore, users no longer have to rely solely on lists of known malicious sites in order to prevent cyber attacks. Machine Learning (ML) is becoming a key aspect of cyber security. It will likely have a major effect on the fight against cyber-crime. But what exactly is it and how can ML secure devices from cyber attacks?

Machine learning is a branch of AI based on the concept that systems can learn from data, identify patterns, and make decisions with minimal human intervention. The idea that computers can learn from previous computations and results to produce reliable and repeatable decisions is not recent. However, growing volumes of available data and powerful developments in computational processing have led to a resurgence and growth in the process. The use of machine learning has already been deployed in various different forms, from Google’s self-driving cars to Netflix’s movie suggestion algorithm, but as technology continues to develop to produce models that can analyze bigger, more complex data and deliver faster, more accurate results even on a very large scale, it is clear that machine learning is fast becoming the new frontier in digital service delivery and has especially become a major factor in the area of cyber security.

Corrata’s SafePathML solution

In response to ‘zero-day’ phishing attacks, Corrata have developed a cutting-edge mobile Machine Learning security solution. Our ML solution can preemptively detect and block phishing attacks in real time. Corrata has identified a range of parameters which can act as indicators of unsafe domains. Using our dataset of malicious and safe domains, we continuously train our SafePathML algorithm to accurately assess the probability of a domain being unsafe. This allows us to block threats even before the wider cyber security community has identified them.

Machine Learning is based on continuous learning and the Corrata solution is constantly improving. As the model successfully identifies phishing attacks and malicious sites, it is refining its accuracy and ability to recognize the parameters of unsafe domains and is therefore constantly increasing its reliability to protect mobile devices from ‘zero-day’ threats. By working with existing threat intelligence databases, Corrata’s SafePathML solution  protects employee devices from all phishing or malicious cyber attacks.

 

To find out more about Corrata’s Machine Learning solution and how it can secure and protect your mobile devices from cyber-attacks, visit www.corrata.com or email us at info@corrata.com.