AI will soon be able to crack even the most advanced security in seconds claims NATO-backed security startup

The cybersecurity world is on high alert as a NATO-supported UK startup warns that artificial intelligence could soon power cyber weapons capable of bypassing even the most sophisticated defences.  

According to a report by Goldilock, such malware may become a reality within two years, leaving companies and governments with little time to prepare. The report highlights the urgency for immediate investment in advanced cybersecurity measures and global policy coordination to safeguard critical systems.

The rise of agentic malware

Goldilock’s report focuses on the looming threat of agentic malware—AI-driven programs capable of infiltrating networks, learning their structures, and evolving to evade detection. The startup compares this future threat to Stuxnet, the infamous worm believed to have been used by the U.S. and Israel to disrupt Iran’s nuclear programme in 2010. While Stuxnet targeted specific systems, agentic malware could autonomously seek out new vulnerabilities and wreak havoc on a global scale.

Critical infrastructure, including energy grids, transportation networks, financial institutions, and healthcare systems, is particularly at risk. The report warns that adversarial governments or cybercriminal groups may deploy such malware to cause widespread panic and destabilisation.

More from Tech

How ChatGPT is becoming everyone’s BFF and why that’s dangerous America ready for self-driving cars, but it has a legal problem

A lack of guardrails

Goldilock co-founder Stephen Kines expressed concern about the rapid pace of AI development without adequate safeguards. The democratisation of AI means that both state actors and cybercriminal gangs can easily access, learn, and exploit these technologies. While some argue that AI-powered security tools could counteract AI-driven malware, Kines cautioned that this approach alone won’t suffice. Instead, a multi-layered defence strategy is necessary, combining AI-enhanced detection systems, network segmentation tools, and innovative solutions like Goldilock’s remote “kill switch,” which can disconnect critical systems in the event of an attack.

Time to act: collaboration and innovation needed

To prepare for the emerging threats, organisations must start investing in advanced cybersecurity tools and fostering real-time collaboration with government agencies and other corporations. Governments, too, need to ramp up investments in AI-driven defences and establish robust policies to regulate AI’s use in cybersecurity. Goldilock emphasises the importance of proactive measures, urging stakeholders to view AI-powered cyber threats as an imminent challenge rather than a distant possibility.

As the global reliance on AI continues to grow, the battle to secure critical systems from AI-driven cyberattacks is becoming a race against time. The message is clear: the clock is ticking, and the world must act now to strengthen its digital defences.