Insights

Artificial intelligence has changed the speed, scale and complexity of cyber risk. Attackers can use AI to automate reconnaissance, discover exposures, develop...

AI is changing cyberattacks. It can accelerate reconnaissance, help attackers analyse potential vulnerabilities, improve phishing campaigns, create convincing synthetic identities and process huge amounts of information considerably faster than a human attacker could alone. But there is an important point at risk of being lost amid all the talk about AI-powered cyber threats.

The European Commission’s Action Plan on Cybersecurity and Artificial Intelligence, published earlier this month, sets out how the EU intends to respond to the threats and opportunities presented by AI.

Vulnerability exploitation has overtaken stolen credentials as the leading initial access vector for breaches for the first time in 19 years. AI is one of the reasons why.

Artificial intelligence has become the defining technology story of the past two years. It dominates boardroom discussions, investor briefings, vendor messaging, product launches and conference agendas. Every sector is being told that AI will transform how it operates, competes and manages risk.Cybersecurity is no exception.

The arrival of Mythos AI has sparked intense debate across the cybersecurity industry. From security researchers to mainstream media, the conversation has swung between fascination and fear to indifference with some stating that we’ve all seen such hype before.

Artificial intelligence is increasingly being embedded across business operations, decision-making processes, and customer interactions. While this presents clear opportunities for efficiency and innovation, it also introduces a new layer of complexity, particularly around governance, security, and compliance.

Organisations across every industry and sector are racing to adopt AI agents. They’re told that they can work faster, automate and reduce costs. Vendors are pushing the idea of intelligent digital workers that can make decisions and complete tasks at pace and with very little human intervention.

Anthropic’s Claude Mythos Preview has triggered a wave of bold claims about the future AI and its role in cybersecurity. Reports of thousands of vulnerabilities discovered, including issues that have supposedly existed undetected for decades, position the model as a transformative force in both attack and defence. Alongside this, initiatives like Project Glasswing signal an industry effort to get ahead of the risks before they fully materialise.

The rapid rise of artificial intelligence across business feels new, urgent and at times overwhelming. Yet for many technology and security leaders, there is a strong sense of déjà vu. The conversations happening today about AI adoption closely mirror those that surrounded early cloud adoption more than a decade ago. The same mixture of excitement, scepticism, regulatory anxiety and skills shortages is resurfacing, just with different technology at the centre. 

Deepfake video, synthetic voice, and AI-generated personas are now capable of mimicking real people with unsettling accuracy. For many organisations, the challenge is no longer spotting a poorly written phishing email, but determining whether the person on the phone, in a video call, or in an email thread is even real.

Our theme for this year’s Security First conferences and our 2026 Trends and Predictions Guide is Resilience redefined: securing the human-AI era. But what exactly does that mean? In this blog we explain it and why it will be the theme of 2026.