COMMENTARY

AI in Cybersecurity: A Framework for CISOs to “Mind the Blade”

Artificial intelligence (AI) is a present-day reality reshaping the cybersecurity landscape. For chief information security officers (CISOs), the integration of AI into security frameworks is a double-edged sword. AI promises enhanced efficiency, predictive capabilities, and automation for internal security teams. Simultaneously, it also endows bad actors with new tools to exploit vulnerabilities across complex ICT supply chains.

AI in Cybersecurity: 2025, the year deployment dwarfs readiness.

AI adoption in IT and cybersecurity has quickly become pervasive. The latest Omdia cybersecurity decision maker survey of nearly 1,000 executives revealed that 93% of organizations are already leveraging AI-powered tools. Further, 91% of firms have scrambled to implement specific AI security policies as a knee-jerk reaction to AI’s ubiquity.

But there are cracks showing. The very tools meant to improve ICT are themselves creating issues. Omdia research found that only 14% of CISOs feel fully prepared to integrate AI into their cybersecurity operations, especially Agentic.

More alarmingly, supporting transformation projects from the chief information officer (CIO) or CEO is the single biggest challenge for more than 50% of CISOs, surpassing challenges from insider threats and ransomware attacks.

Data privacy and identity security are the most significant concerns, accounting for nearly 70% of CISOs’ worries globally. Ethical governance gaps, a shortage of skilled personnel, and budget constraints further complicate AI adoption. Additionally, many organizations struggle with knowing where and how to start, highlighting the need for clear frameworks and actionable road maps.

AI is not a single technology or model: The nuance is important.

Current deployments of AI in ICT and cybersecurity range from longstanding, robust machine learning to generative AI (large language models), and more recently, agentic AI (semi-autonomous). Each class of AI has different strengths and limitations across threat detection, response automation, and data analysis.

CISOs need to reframe where and how AI is used across complex technology supply chains, including third parties, in line with evolving regulations, expanding threats, and more sophisticated attacks. Over 70% of firms surveyed have more than 20 standalone cybersecurity tools in use. Twenty percent have more than 50. Each tool will have some level of maturity across AI as an inbuilt capability or a data source for another tool.

The challenge is immense. As a starting approach, Omdia research from decision makers reveals five frames of reference that must be applied across ICT, AI, and Cybersecurity as an integrated approach.

Omdia’s Five Dimensions of AI in Cybersecurity

To effectively harness AI, CISOs must consider its role across five critical dimensions:

Cybersecurity with AI (augmentation): AI can augment existing security operations, enabling analysts to access vast repositories of knowledge and identify threats more efficiently. Generative AI, for instance, can assist in sorting and analyzing data from tools like TDIR (threat detection and incident response), SIEM (security information and event management), and SOAR (security orchestration, automation, and response). However, it is crucial to maintain a “human-in-the-loop” approach to ensure AI outcomes are interpreted correctly and complemented by human judgment. 1.

Cybersecurity by AI (automation): AI has the potential to independently deliver cybersecurity capabilities, such as faster threat detection and response. This dimension has been evolving since the emergence of machine learning in cybersecurity around 2014/15, particularly with the rise of EDR (endpoint detection and response). Predictive security, powered by specialized AI models, can help organizations stay ahead of threats. However, automation must be balanced with transparency to build trust in AI systems. 1.

Cybersecurity for AI (tooling): As AI becomes more pervasive, protecting the AI systems themselves is paramount. Attackers are increasingly targeting algorithms and training data to manipulate outcomes. CISOs must invest in security tooling that safeguards AI models and accelerates incident response. The rapid adoption of AI-enabled applications demands a reframing of security strategies to keep pace with evolving threats. 1.

Cybersecurity against AI (defense): AI is a double-edged sword. While it can be used to enhance security, it is also exploited by attackers to create more sophisticated threats. For example, AI-powered deepfakes and automated DDoS attacks pose significant challenges. CISOs must deploy advanced detection technologies, often leveraging AI, to counter these threats effectively. 1.

Cybersecurity and AI (strategy and governance): AI strategies must align with business objectives to ensure innovation does not compromise security. CISOs are under pressure to exploit AI for commercial gains while navigating global economic uncertainties and competitive pressures. Developing a comprehensive AI cyber strategy that integrates governance, ethical considerations, and risk management is essential.

Recommendations for CISOs to Grapple and Harness AI in Cyber

This space is evolving, and Omdia continues to track, evaluate, and test vendor capabilities across endpoint, cloud, network, data, identity, and SecOps tools where “AI” is offered. There is a lot of AI washing and misrepresentations of capabilities today, not surprisingly, given the rate of market change and technology innovation.

Omdia recommends that CISOs navigate the complexities of AI in cybersecurity in five areas:

Prioritize readiness: Invest in training and infrastructure to support AI adoption. Building in-house expertise and fostering employee readiness are critical to bridging the readiness gap. 1.

Focus on governance: Develop ethical frameworks and governance tools to ensure AI systems are transparent, trustworthy, and compliant with regulations. 1.

Adopt predictive security: Leverage AI for threat intelligence, exposure management, and automated remediation. Predictive security can help organizations stay ahead of emerging threats. 1.

Protect AI systems: Implement robust security measures to safeguard AI models, algorithms, and training data from attacks. 1.

Align AI strategies with business goals: Ensure AI initiatives are clearly linked to business objectives to balance innovation and security.

Final Word

The stakes have never been higher. As AI pops up in more places, including mission-critical enterprise platforms such as Salesforce, Oracle, SAP, and Microsoft, CISOs must act decisively to harness AI’s potential while mitigating its risks. This requires a balanced approach that combines technological innovation with human expertise and ethical governance. Neglecting any of the five dimensions of AI in cybersecurity will result in missed opportunities and increased vulnerabilities.

Related Links:

Omdia Cybersecurity

Omdia Analyst Summit at Black Hat USA 2025

Omdia Data and Identity Security Intelligence Service

About the Authors

Practice Leader, Cybersecurity, Omdia

Adam is part of Omdia’s global Digital Enterprise Services practice. He leads international research, advisory, and analysis for cybersecurity and cloud B2B services. A regular speaker at industry events and executive roundtables, Adam thrives on driving deep insight through custom engagements with clients. Adam joined Omdia in 2020, bringing more than two decades of experience as an executive at some of the world’s largest telecoms and IT services vendors. His experience spans corporate strategy, product marketing, technical operations, market intelligence, sales, and strategic alliance roles. Adam holds a Bachelor of Economics from Macquarie University and an MBA from Deakin University (both in Australia). In 2022, he was awarded the 2021 Brookes Medal for outstanding academic achievement.

Senior Principal Analyst, Cybersecurity, Omdia

Rik is senior principal analyst in Omdia’s IT security and technology team, specializing in cybersecurity technology trends, IT security, compliance, and call recording. He provides analysis and insight on market evolution and helps end users determine what type of technology and which vendor they should be pursuing.

Rik has also worked in Omdia’s financial services technology team, with a specialization in capital markets technology. Prior to joining Omdia, he worked as an IT journalist, specializing in networking and security, and as a foreign correspondent in Brazil, where he worked, among others, for the Financial Times and The Economist.