Phishing remains one of the top threats to organizations. According to Proofpoint’s 2019 State of the Phish report, 83% of organizations experienced phishing attacks. While most inboxes come with basic protection like spam filters, it’s often not enough. A 2019 Global Phish Report found that 25 per cent of phishing emails bypass default security. To defend against the ever-evolving threat of phishing, many email security solutions are integrating artificial intelligence (AI). 

Source: Statista

Using AI To Fight Phishing 

Automated Detection

AI goes beyond signature-based detection. Signature based detection analyzes attack patterns to determine incoming attacks. A phishing signature will include information like known malicious IPs or domains or specific types of email headers. If a signature is identified, the email will automatically be classified as phishing and will be blocked. Signature based detection is important, especially for detecting known attacks. However, hackers know that by tweaking certain elements like HTML code or image metadata, they can evade these filters. 

AI enhances detection with machine learning capabilities. AI scans for similar signature features such as malicious links but has other abilities like scanning images to determine fake login pages. AI also focuses on detecting phishing characteristics/behaviours as opposed to known signatures. Even if a hacker sends a phishing email with altered signature, AI will be able to automatically detect it as a phishing email and block it accordingly.

Behavioural Analysis

Some phishing attacks like CEO fraud and social engineering imitate users or companies to trick victims into wiring money or providing confidential information. These attacks can bypass spam filters because they come with no malicious payloads (such as links or attachments). A user who receives these kinds of emails may think they are legitimate, especially if they look very convincing.

AI can analyze user communication patterns to establish a baseline for normal user behaviour. This makes it easier to detect anomalies if hackers are trying to impersonate employees. Behavioural analysis can examine characteristics like writing style and grammar and syntax to determine a user’s unique profile. If a user receives an email claiming to be from the CEO but the message has grammatical errors it can be flagged as suspicious, even if the headers and sender address look legitimate. 

Continuous Learning 

Phishing attacks are always evolving, and threat actors are always trying to evade the latest security technologies. In order to stay ahead of threats, cybersecurity technologies also have to evolve and improve. 

Artificial intelligence is continuously learning from not only an organization’s unique environment but also from other open source threat intelligence feeds. By constantly updating and feeding its capabilities, AI is able to constantly improve its ability to detect the latest phishing threats.   

Protect Against Phishing

A combination of trained users and using evolving tech like AI is necessary for organizations to defend against phishing attacks. At Jolera, our Secure IT – Mail solution provides comprehensive security, archiving, backup and user defence features to protect your organization. Included in the solution is an AI component that protects email against spear phishing. For more information on how Secure IT – Mail can protect your organization, contact us today.