36 Group Unveils Trustplay aOS: The World's First AI Agent-Native Operating System for Online Casinos
KUNGSBACKA, SWEDEN — 13 APRIL 2026 — 36 Group AB today launched Trustplay aOS (Agentic Operating System), the online casino industry's first operating system where AI agents and human operators share the same state, memory, and compliance rules. Trustplay aOS represents the evolution of 36 Group's PAM business and will serve as the operating system for the group's own consumer casino brands.
For decades, online casinos have scaled by putting hundreds of people in front of monitors, paying them to click buttons, approve withdrawals, and read reports. Current attempts to bring AI into the back-office try to teach machines to navigate software designed for humans. Trustplay aOS takes the opposite approach: the platform itself is the agent interface, and humans supervise from a dedicated control room.
"Bolting AI onto a casino back-office built for humans is a waste of time, and most AI in iGaming is just theatre," said Chris Steele, CEO of 36 Group. "Trustplay aOS is infrastructure. The agents do the work, the humans set the rules, and nothing happens that the system can't explain afterwards."
The Hive Mind
Most AI deployments rely on isolated bots that forget context the moment a session ends. Trustplay aOS replaces them with a centralized, persistent intelligence shared across every agent in the system. When one agent learns something, from new fraud patterns to a shift in player behavior, every agent who benefits from this new knowledge has access to it instantly. Intelligence is centralized, persistent, and compounding.
Compliance by architecture
Agents operate inside a strict compliance boundary that prevents unauthorized actions. High-risk decisions are routed to human operators for one-click approval. Routine work runs autonomously. Material decisions stay with people.
Proactive operations
Trustplay aOS does not wait to be asked. When a meaningful event occurs anywhere in the casino, the system gathers context, analyzes risk, and stages a recommendation before a human operator opens the ticket. It runs proactively on schedules and reactively on live data.
Continuous learning, human-approved
When agents encounter outdated procedures or friction in the rulebook, they cannot rewrite the rules themselves. They surface a proposed change to human operators. Once approved, the new procedure propagates across the system. The platform learns continuously. Only humans change the rules.
The Control Room
Every agent action is observable from a dedicated command center in the Trustplay dashboard. Operators can chat with agents in real time, audit the reasoning behind any decision, and correct agent behavior on the fly.
Trustplay aOS is engineered as production infrastructure for licensed operators. By decoupling operational growth from human headcount, it drives the marginal cost of scaling a casino toward zero.
Trustplay aOS is available via 36 Group's Trustplay division. Contact sales@36.group for partnership and licensing enquiries.