The evolution of Cybersecurity and AI

The evolution of Cybersecurity and AI

Over the past 12 months, there has been an evolution in the role of Generative Artificial Intelligence (GenAI) in cybersecurity settings wherein organisations are no longer considering if they will utilise the benefits of GenAI to help with Cybersecurity, but rather how these benefits will be utilised.

2024 in particular has seen an increase in investment in GenAI, with a mix of risk management in tandem with new capabilities being introduced as a result of AI. According to a recent survey by PwC, 74% of organisations have increased their investment in this period, to take advantage of the new technology. Yet 66% of security leaders feel their risk surface has increased as a consequence. So how then, do we ensure that AI is adapted in a way that minimises Cybersecurity risk?

There’s plenty of debate about the use of AI in Cyber defences, perhaps more from a perspective that because the bad-actors are utilizing this new approach, that the good-actors are compelled to do similarly. But if we look at why AI is to be used for Cyber defence, the story is somewhat different. The evolution of threats is real, yet the evolution of Cyber defence is also a controllable aspect using AI. The adaption of constant learning, scanning and updating of defence systems that are provided by alliance partners or developed in-house allows an organisation to monitor its threat environment in real-time. This intel can used by a security team for real-time alerts. No longer are we faced with a reactive stance on threats, now we have proactive threat hunting in real-time, that, if the LLM is structured to do so, will allow for an evolution of learning and adapting to threats continuously.

Why not pivot to using GenAI as a component of your security and risk management? Although GenAI is increasing the cyber risk attack surface for most organisations, organisations can also using that same technology for cyber defence. For example, AI would be a useful tool for threat detection and response, threat intelligence and malware/phishing detection. To introduce such an approach also means overcoming potential internal stakeholder mistrust of GenAI, Inadequate internal controls and risk management in relation to the use of GenAI and a lack of standardised internal policies governing the use of GenAI
These are surmountable issues, and the used of GenAI in Cybersecurity can provide a secure platform to allow GenAI to be used in more and more organizations, as risks can be monitored in real time and threats can be detected with an ever-evolving security model.

Statistics are taken from PwC’s 2024 DTI report (https://www.pwc.com/dti)

Author: Neil Redmond, Director Cyber Security, PWC and graduate of the Executive MBA