For the fourth time, Mitiga has been named one of the 30 best startups to work for in Dun & Bradstreet’s 2026 ranking of companies with up to 100 employees in Israel, but this one means quite a bit more than the others. Why? Because D&B also published its first list of startups leading in AI adoption and implementation this year, and Mitiga is one of the 5 on it, with Loox, Komodor, and us as the only 3 startups on both lists.
Why did Mitiga make the list?
The people who work at Mitiga did most of the work, since D&B's questionnaire answers belong to them, and a full 80% of the team filled out D&B's employee survey.
"People choose Mitiga to work alongside talented colleagues, tackle complex cybersecurity challenges, and have the trust and ownership to make a real impact," says Rony Sarig, our VP of HR.
From what I can see in my chair with marketing, the Mitiga team moves fast and shares the load, and both of these are foundational values for the company. Our customers face attackers who move at machine speed, and that sets the pace inside the company, too.
What does leading in AI adoption look like at Mitiga?
D&B's AI evaluation looked at our AI transformation and what it has changed for our product, our customers, and our employees. Its team met with Ofer Maor, our co-founder and CTO, who walked them through how we did it.
"There's a huge difference between adopting and using AI versus becoming AI native," Ofer says. "The latter is a huge transformation."
In a little over a quarter, we rebuilt how Mitiga works around AI. We mixed training and experimentation and show and tell with an AI revolution kick-off. The most difficult part was to get people to challenge processes that they have built and used for years.
This AI sprint period gave teams dedicated time to break those processes and rebuild them. And it worked. In an employee survey, our people gave AI in their daily work a 5.0 out of 5 and our AI training a 4.8, compared to an industry average of 4.5. People worked across team boundaries and turned expert knowledge into reusable skills and workflows.
Some of what changed:
- Our R&D team built a multi-agent system that drafts detection rules, tests them adversarially, and checks them against real customer data, with people reviewing the results. It moved us from 50 detections a month to more than 1,000, while freeing up team time.
- We reached more than 85% agentic pull requests.
- Supporting a new data source went from half a quarter to half a sprint.
- A fully agentic requirements-to-production workflow now delivers new capabilities in days.
Mitiga Labs also ran a Hugging Face-style agentic attack against our environment to see what our detections would catch, and we joined Anthropic's Cyber Verification Program.
Mitiga's Helios AIDR is built with AI to defend with it, from it, and the AI infrastructure itself.
What makes Mitiga different as a place to work?
Mitiga AI Agent Security is built on one normalized Cloud Security Data Lake that spans cloud, SaaS, identity, and AI. It answers what an agent did, how far it reached, and what to do right now. Our mission is Zero-Impact Breach Prevention.
"We look for curiosity, ownership, and the ability to turn seemingly impossible ideas into action," Rony says. "People need to bring depth in their field while staying flexible, learning quickly, and working closely with others."
You can learn more about Mitiga on our company page.
