Cyber vulnerabilities found by internal research teams up 80%
Cyber-physical system vulnerabilities disclosed in the second half (2H) of 2022 have declined by 14% since hitting a peak during 2H 2021, while vulnerabilities found by internal research and product security teams have increased by 80% over the same time period, according to the State of XIoT Security Report: 2H 2022 released today by Claroty.
These findings indicate that security researchers are having a positive impact on strengthening the security of the Extended Internet of Things (XIoT), a vast network of cyber-physical systems across industrial, healthcare, and commercial environments, and that XIoT vendors are dedicating more resources to examining the security and safety of their products than ever before, the report states.
Compiled by Team82, Claroty's research team, the sixth biannual State of XIoT Security Report is a deep examination and analysis of vulnerabilities impacting the XIoT, including operational technology and industrial control systems (OT/ICS), Internet of Medical Things (IoMT), building management systems, and enterprise IoT.
The data set comprises vulnerabilities publicly disclosed in 2H 2022 by Team82 and from trusted open sources including the National Vulnerability Database (NVD), the Industrial Control Systems Cyber Emergency Response Team (ICS-CERT), CERT@VDE, MITRE, and industrial automation vendors Schneider Electric and Siemens.
Amir Preminger, VP research at Claroty, comments, "Cyber-physical systems power our way of life. The water we drink, the energy that heats our homes, the medical care we receive - all of these rely on computer code and have a direct link to real-world outcomes.
"The purpose of Team82's research and compiling this report is to give decision makers in these critical sectors the information they need to properly assess, prioritise, and address risks to their connected environments, so it is very heartening that we are beginning to see the fruits of vendors' and researchers' labor in the steadily growing number of disclosures sourced by internal teams.
"This shows that vendors are embracing the need to secure cyber-physical systems by dedicating time, people, and money to not only patching software and firmware vulnerabilities, but also to product security teams overall."
Key findings include the following:
- Affected devices: 62% of published OT vulnerabilities affect devices at Level 3 of the Purdue Model for ICS. These devices manage production workflows and can be key crossover points between IT and OT networks, thus very attractive to threat actors aiming to disrupt industrial operations.
- Severity: 71% of vulnerabilities were assessed a CVSS v3 score of "critical" (9.0-10) or "high" (7.0-8.9), reflecting security researchers' tendency to focus on identifying vulnerabilities with the greatest potential impact in order to maximise harm reduction. Additionally, four of the top five Common Weakness Enumerations (CWEs) in the dataset are also in the top five of MITRE's 2022 CWE Top 25 Most Dangerous Software Weaknesses, which can be relatively simple to exploit and enable adversaries to disrupt system availability and service delivery.
- Attack vector: 63% of vulnerabilities are remotely exploitable over the network, meaning a threat actor does not require local, adjacent, or physical access to the affected device in order to exploit the vulnerability.
- Impacts: The leading potential impact is unauthorised remote code or command execution (prevalent in 54% of vulnerabilities), followed by denial-of-service conditions (crash, exit, or restart) at 43%.
- Mitigations: The top mitigation step is network segmentation (recommended in 29% of vulnerability disclosures), followed by secure remote access (26%) and ransomware, phishing, and spam protection (22%).
- Team82 contributions: Team82 has maintained a prolific, years-long leadership position in OT vulnerability research with 65 vulnerability disclosures in 2H 2022, 30 of which were assessed a CVSS v3 score of 9.5 or higher, and over 400 vulnerabilities to date.