AI is forcing companies into environments where they are no longer competing against rivals, but with adversaries. In cybersecurity, where attackers adapt in real time, the competitive advantage for defenders is no longer just about efficiency, but in how companies learning and revise their assumptions.

Instead of optimizing for speed alone, some of the biggest AI security companies are being built around principles closer to IDF intelligence operations that include constant simulation, adversarial thinking, after-action reviews, and learning under uncertainty.

To better understand this, I spoke to Ido Geffen, Co-Founder and CEO at Novee Security. Novee is an AI-driven cybersecurity startup focused on transforming how companies find and fix security vulnerabilities.

Rather than traditional penetration testing, which is manual, periodic, and slow, the company uses continuous, autonomous AI that simulates real-world attackers to proactively uncover hidden vulnerabilities and validate them in real time.

Geffen described modern cyber defense as fundamentally imbalanced. Defenders must be perfect all of the time, whereas attackers only need to succeed once to penetrate a company. Its platform blends offensive cyber tradecraft with AI to close the gap between evolving threats and slower defensive testing processes.

“You can be perfect in 99.9% of the time, but then you have this one hole, and you lose the game,” he explained. “And the bad guy doesn't have these constraints… they just need one shot that will work.”

The new reality facing companies and cyber defense technologies means there is a shift in how to tackle these threats. Traditional enterprise software ships features, but companies like Novee continuously discover failure. This is where the mindset shifts from an engineering discipline to an intelligence discipline.

Geffen, who has experience in the IDF, including time spent in Gaza’s dangerous tunnel system, explained how the differences between preparedness and adaptability were a military skill he adopted and brought to the cybersecurity sector.

“One of the main things that I learned during those years is the ability of every soldier or a man in this type of unit to come up with ideas on how we can improve, what we can do differently,” he explained. “So I think this is one of the strengths of this type of unit. It’s hierarchy and all of that, but in the end, if someone comes up with a very good initiative, you can really change the way things are being done. And I think that, at least for me, it was very inspiring.”

Thanks for reading The Spiro Circle! This post is public, so feel free to share it.

The AI era is creating companies that operate in new environments where the problem space evolves daily - and cybersecurity companies are forced to keep up with the hackers. In those conditions, efficiency is key, and adaptability must operate alongside preparedness.

The company emerged from stealth last month with a $51.5 million funding raise across Seed and Series A in just four months since its founding in May 2025 - a huge testament to its tech, talent, and market position. Backers include YL Ventures, Canaan Partners, and Oren Zeev via Zeev Ventures.

As AI agents interact with complex real-world systems across finance, infrastructure, or healthcare industries, the companies building them will need to adopt strategies that resemble military intelligence: constantly testing reality, updating beliefs, and assuming uncertainty is permanent.

Get full access to The Spiro Circle at www.thespirocircle.com/subscribe