Cyberattacks increased by 40% in 2021, and Italy was the 4th most affected Country by Covid-19-themed cyber threats in the first six months. As absurd as it may sound, this situation makes us clearly understand how protecting ourselves only from what has already been classified as a threat isn’t enough, and how threats that aim to steal our data have evolved. Let’s talk, in particular about phishing.                                                           

Phishing is a type of scam through which an attacker tries to trick the victim into providing personal information, financial data or access codes, pretending to be a trustworthy entity in a digital communication. The phishing attempt doesn’t happen anymore only through suspicious e-mails popping in our inbox, but nowadays fits perfectly the victim’s profile: cyber criminals can easily draw up a profile of us because of the numerous extensions and applications that trace our online behavior. Every day hundreds of thousand new phishing sites are created.

A special kind of phishing attack is the short-lived phishing, often programmed to have extremely short life cycle able to cross common security solutions: their life cycle is a few days or even hours, thus failing to be detected as malicious by traditional solutions. Short-lived phishing success rate is high as they are actual zero day – in fact, a Zero-Day is a software vulnerability known only by those who discovered it and for which there are no remedies or protections yet. In the specific case of phishing, a Zero-Day is the phishing website that has not yet been discovered.

A typical phishing website life-cycle.

Considering the difficulties of detection and defense, it’s not surprising that Zero-Day exploits have grown to become one of the most popular methods chosen by hackers. Not surprisingly, many of the most serious cybersecurity breaches globally are Zero-Day attacks, and recent studies estimate that about 30% of all malware attacks target Zero-Day vulnerabilities.

Because only a small percentage of vulnerabilities are discovered (and it always happens when it’s too late), this is clearly an important potential growth area for cybercrime – just look at some of the most recent Zero-Day attacks. Cybercriminals’ tactics are always evolving, and while protection from known threats requires traditional solutions such as good quality antivirus, antispam and firewalls, Zero-Day attacks require the support of Artificial Intelligence.

In this regard, the Ermes – Intelligent Web Protection R&D Team has recently developed a solution to prevent Zero-Day attacks that is extremely competitive compared to all others on the market: Ermes 0Day Protection. The innovative algorithm, in fact, identifies phishing sites with a 48/72 hours of advantage as soon as they are born, while other solutions typically rely on reports from users or, even worse, require the user to click on the malicious link before taking action, putting their security at risk.

Ermes 0Day Protection builds on Ensemble Learning techniques which consider hundreds of features to quickly identify phishing pages. It has been designed to be fast, scalable, yet precise, so that it can process millions of domains per day while guaranteeing high accuracy at detecting web design patterns typical of malicious webpages.