What’s on my mind in cybersecurity right now

Global, Jun 22, 2026

What’s on my mind in cybersecurity right now 

By Martie Moore 

As a cybersecurity professional with more than 25 years’ experience, I have always known security to be fast-moving. But the pace of change being driven by AI feels different. This is not just another evolution in technology; it is a fundamental shift in how threats emerge, how quickly they scale and how organisations need to respond. 

When I was asked to introduce the June issue of Cyber Bytes, three topics stood out. Each of them ultimately came back to AI. So, in this issue, I want to share three converging security themes that are top of mind for me right now. 

 

1. Shadow AI is exposing the governance gap 

Governance is no longer just about data, it is about how AI is used, deployed, monitored and controlled across the organisation. 

This is where shadow AI is creating pressure. Employees are adopting AI tools at speed, often outside approved environments. The Logicalis Global CIO Report 2026 found that only 37% of CIOs have full visibility of the AI tools in use across their organisation, while 41% say employees are already putting data security at risk through how they use AI tools. 

Many organisations are still defining the policies, guardrails and accountability needed to manage AI use responsibly. Detection capabilities may exist, but the real challenge is governance maturity across the entire AI lifecycle. Without it, shadow AI will continue to outpace formal security frameworks. 

 

2. Vulnerability management is being redefined by AI-scale threats 

For a long time, vulnerability management followed a familiar sequence: find the vulnerability, assess the risk, patch before an attacker gets there. It was not perfect, but it was predictable, and there was time. 

That time is disappearing. AI can analyse vast codebases and uncover flaws at a scale and speed that was not previously possible, including vulnerabilities that may have sat undiscovered for years. 

The bigger disruption is what happens next. AI systems can validate findings, test them and combine multiple weaknesses into viable exploit paths, compressing the gap between discovery and exploitation. 

That fundamentally changes the problem. Discovery is no longer the bottleneck, response and remediation is. Security teams now need to prioritise and remediate fast enough before exposure turns into impact. 

Project Glasswing gives a glimpse of what comes next. Anthropic describes it as an initiative to secure critical software by giving selected organisations early access to Claude Mythos Preview for defensive security work. Launch partners include Microsoft, Cisco, Google, AWS and others, with the aim of identifying, testing and remediating vulnerabilities at scale before they can be exploited. 

What makes this significant is the direction of travel. AI-led security is starting to move beyond individual environments and towards vendor ecosystems, where coordinated, large-scale patching and response could become possible. 

But there is also an asymmetry emerging: vulnerabilities can be uncovered faster, exploit paths can be assembled faster, and access to the most advanced capabilities is not evenly distributed. 

The question for security leaders is no longer simply how fast they can find risk. It is whether they can build the operating model, automation and ecosystem coordination needed to remediate at AI speed. 

 

3. Securing AI is still in its infancy 

Perhaps the most important, and least mature area, is securing AI itself. 

New threat vectors are emerging rapidly, from data poisoning through to model reliability issues and misconfiguration. At the same time, organisations are racing to deploy AI use cases, often without fully validating security, governance or output quality.  

In many cases, this is not about negligence, it is about maturity. Skills, frameworks and processes are still catching up with the pace of adoption. 

Right now, many organisations are building AI capability faster than they can secure it.  AI is creating both opportunity and instability in equal measure. With breakthroughs like Glasswing and the evolving AI threat landscape, the rules of cybersecurity are being rewritten in real time. 

Topic

Related Insights