DOMAIN / 01ACTIVE
Artificial Intelligence
Reasoning, learning, and adaptive computation for systems that operate beyond the lab.
Explore domain →Independent research organization
Advancing intelligence.
Securing what comes next.
Gtomac works across artificial intelligence, cybersecurity, autonomous systems, and emerging computational security.
SCROLL TO EXPLORE ↓01Research domains
Difficult problems. Connected disciplines.
DOMAIN / 02ACTIVE
DOMAIN / 03ACTIVE
DOMAIN / 04ACTIVE
02Featured research
Work at the edge of capability and control.
FIELD / AI SECURITY
AI Agent Security
Evaluating autonomous systems under adversarial conditions.
→FIELD / INFRASTRUCTURE
Secure Model Infrastructure
Building resilient foundations for frontier AI systems.
→FIELD / MACHINE INTELLIGENCE
Adaptive Computation
Research into reasoning, learning, and adaptive systems.
→03GTOMAC RESEARCH NETWORK
One mission.
Six connected fronts.
Breakthroughs rarely respect discipline boundaries. Our work connects intelligence, systems, and security into one continuous research practice.
ACTIVE DOMAINCROSS-DOMAIN LINK
GTOMACGCORE / 00
AI01
AGENTS02
SAFETY03
SECURITY04
SYSTEMS05
CYBER06
04Latest publications
Ideas, evidence, and open questions.
View all publications →GTOMAC / OPEN-RESEARCHMAIN ⑂
⌄ ▣ gtomac/
⌄ ▣ research/
⌁ artificial-intelligence/
◇ cyber-security/
◉ agent-security/
› ▣ datasets/
› ▣ benchmarks/
› ▣ open-source/
≡ README.md
9f42c7aUpdate agent boundary benchmark4h ago
05OPEN RESEARCH
Research compounds when it is shared.
We publish select code, datasets, benchmarks, and experimental methods for the wider research community.
Explore open projects →06JOIN THE LAB
Work on difficult
problems that matter.
We bring together researchers, engineers, and security practitioners who want to shape safer computational systems.
View open roles →