ZioSec Team
Superhero Capital is excited to announce its investment in ZioSec, a US-based cybersecurity startup building the first offensive security platform designed to protect AI agents and agent-driven workflows in production environments.
As companies race to adopt AI, security tools are struggling to keep up. ZioSec steps into this gap with a bold mission: break AI systems before bad actors do.
Founded by Aaron Walls and Andrius Useckas, ZioSec is taking a new approach to AI security by simulating real-world attacks on AI systems—testing their resilience through aggressive, continuous offensive security techniques. From prompt injection vulnerabilities to broken agent control paths, ZioSec helps companies identify and address weaknesses before they make headlines.
$2.1M in Seed Funding Backed by a Strong Network
ZioSec has raised $2.1 million in seed funding, backed by Superhero Capital, alongside a strong syndicate of investors including Access Venture Partners’ Frank Mendicino, Jason Calacanis (LAUNCH), Greater Colorado Venture Fund, Kickstart, and a network of seasoned angels.
This funding will support ZioSec’s product development and go-to-market efforts as it helps enterprise and high-risk industries build more secure AI systems.
Offensive Security, Purpose-Built for AI
Unlike traditional defensive tools, ZioSec’s platform is built from the ground up to address the unique risks of AI agents in production. Its offensive testing framework uncovers vulnerabilities by simulating the types of attacks real-world adversaries are already experimenting with—from adversarial prompts to logic manipulation across agent workflows.
By continuously pressure-testing systems, ZioSec helps organizations ship AI with confidence—without compromising on security.
A Veteran Team Tackling a Frontier Problem
Co-founders Aaron Walls and Andrius Useckas bring deep expertise in cybersecurity, red teaming, and infrastructure defense. Their experience spans decades of building and breaking critical systems—and they’re now channeling that knowledge into building a new category of AI-native security tooling.