Cybersecurity and AI: Tools, Opportunities, and Threats

Keywords: Industrial AI, Cybersecurity, Machine Learning, Deep Learning, Asset Discovery

Overview

AI affects many aspects of industrial cybersecurity, and all end users should be cognizant of these. AI is being used in a growing range of industrial cybersecurity products and applications, but it’s also being used by attackers to mount more effective campaigns and create malware. Meanwhile, end users continue to implement AI technology in many forms, from machine learning applications to large language modeling.

End users need to consider not only how they are going to use AI in their own organizations, but also how they will keep this new technology secure. AI is becoming ubiquitous, with AI functions being embedded into a huger range of software applications and managed services. Cybersecurity suppliers are offering an increasingly wide array of solutions for end users to manage AI within their enterprises, and AI is also making its way into many OT cybersecurity supplier products and applications.

Many end users are not even aware of the scope of AI tools being used at their own company, and their AI adoption strategies can lack the “guardrails” necessary to reduce risk and maintain a resilient security posture. This includes both plant and enterprise operations as well as internal software development activities. Many end users in a wide range of industries from pulp and paper to oil and gas are developing their own large language models and copilots to be used by their own workers. AI is also being used by attackers to mount more effective malware campaigns, identify and exploit vulnerabilities, and create more realistic and deceptive phishing campaigns.

What Is Industrial AI?

Before we continue, a basic definition of Industrial AI is required. In the broadest sense, AI falls under four primary groups: Causal AI, Machine Learning, Deep Learning, and Large Language Models (LLMs). Causal AI focuses on cause and effect relationship in data and allows for more in-depth root cause analysts as to why certain things are happening.

Machine Learning enables systems to learn and improve performance over time through the analysis of large amounts of data without the need for explicit programming. Deep Learning can be considered a subset of Machine Learning and covers areas like speech recognition, image recognition, and language processing. LLMs are intensively trained on large datasets and can generate text, images, audio, or video based on the type of model and the user requests. LLMs are the foundation for things like industrial copilots due to their ability to provide instant replies to queries.

How AI Impacts Industrial Cybersecurity

The relationship between AI and industrial cybersecurity has become quite complex. AI in the form of Large Language Models, Machine Learning, Deep Learning, copilots, and other tools are being increasingly implemented by major end user companies. Many of these AI-based solutions are developed internally. Some companies, for example, are creating Large Language Models as a repository of process knowledge, where workers can ask just about any question regarding any manufacturing process within the company and get an answer instantly.

ARC Advisory Group clients can view the complete report at the ARC Client Portal.

Contact Us if you would like to speak with the author.

Obtain more ARC In-depth Research Market Analysis.