Horizon3 opens lab for closed-loop attack-and-defense AI
Original title: Horizon3 Launches Adversarial Intelligence Frontier Lab
Purdue’s Z. Berkay Celik will lead work that uses live NodeZero attack-path evidence.
Horizon3 said on November 12, 2026, that it launched the Adversarial Intelligence Frontier Lab, directed by Purdue associate professor Z. Berkay Celik. The lab plans production-safe closed-loop systems that use verified NodeZero attack graphs to find and close attack paths, plus work on autonomy controls and critical-infrastructure resilience centers. Horizon3 says the effort draws on tests from more than 7,000 customers.
Unconfirmed: The 30 billion tokens-per-day figure is Horizon3’s own estimate.
- Initial work covers AI attacks, defense learning loops, autonomy safety, and cyber resilience.
- Horizon3 estimates 30 billion anonymized attack-defend training tokens per day.
- The lab will also oversee Cyber Resilience Centers of Excellence for critical infrastructure.
- Celik co-directs the Purdue Security Laboratory.
- Vendors
- Horizon3
- Products
- NodeZero
- Organizations
- Purdue UniversityPurdue Security Laboratory
From the original article
4 min read at horizon3.ai
Horizon3 launches the Adversarial Intelligence Frontier Lab, led by Purdue's Dr. Z. Berkay Celik, to research production-safe closed-loop systems that find and close real attack paths at machine speed.
By Horizon3 · horizon3.ai
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.