2 min

Akamai introduces the AI-based Page Integrity Manager tool to help companies discover malicious code on their websites. The tool also detects vulnerabilities in used solutions from other players, such as payment services.

According to the Content Delivery Network (CDN) specialist, the increase in online shopping is causing companies’ websites to fall victim to web skimming campaigns and other cyber-attacks more often. In particular, attacks in which malicious scripts are used to steal credit card data.

Page Integrity Manager

Akamai wants to protect companies against mostly script-based attacks with the recently launched Page Integrity Manager tool. This tool uses artificial intelligence to detect malicious code on websites that are attempting to steal user data. The tool runs in the browser environment and analyses sessions for suspicious activity. This is how malicious code can be discovered.

More specifically, the tool analyses whether a particular script is malicious by looking at a combination of factors. For example, the AI models check the web address from which a script is loaded in the end-users’ browser. They also check the servers to which the script is sending the data retrieved from the visitors. In addition, the tool evaluates other details such as the number of end-users that correspond to suspicious code.

Dashboard functionality

A dashboard provides administrators with a clear overview of all suspicions. This dashboard gives them an overview of all scripting activities on their website. In addition, they can use specific security tools via the dashboard to limit the risk of malicious scripts, such as imposing certain restrictions on access to end-user data.

Monitoring third-party scripts

According to Akamai, Page Integrity Manager detects threats both in customers’ own scripts and in third-party scripts of external services, such as payment platforms. The tool detects whether the websites of external services have specific software vulnerabilities that are registered in the CVE database.