Blog

In the year 2022 and as is reflected in the article “Industrial Security 2022 in numbers”, cyberattacks in all industrial sectors have increased by around 30 % in the third quarter of 2022 and it is estimated that the number of organizations or industrial manufacturers victims of a cyberattack was around 40% in the last year. Especially in the industrial sector, the number of attacks has grown exponentially due to the massive introduction of IoT devices (it is expected to go from 13.5 to 21.5 million connected devices in three years) or more specifically about IIoT devices, which have been the main gateway for attacks as manufacturers have prioritized features and mass-production of devices over the security. In addition, this is compounded by planned obsolescence planned (increasingly present in this type of devices), increased interoperability and connectivity and the appearance of new types of malware and exploits which are much more effective.

In recent years, the concept of machine learning has gained more prominence, mainly driven by advances in parallel computing capacity. More and more developments, applications and programs are using these algorithms to provide systems with greater security, intelligence and independence. However, it’s rarely used in industrial environments, although some recent tests and developments prove its effectiveness, including in the scope of detection and prediction of cyber attacks.

The continuity of the production process in businesses that require industrial automation depends more and more on the proper functioning, safety and reliability of the system of that composes it. Therefore, conducting tests of acceptance of its operation prior to its commissioning, is vital to ensure that the systems acquired meet the requirements set out in the contract between the company and the manufacturer.

This post will analyse the vulnerabilities associated with Log4Shell, detected in the library Log4j, which is found in infinite software products both in technical and industrial fields. Although there have been other instances of more sophisticated vulnerabilities, the problem with this one is area of exposure.