Product · Safety
Built for regulated markets, not retrofitted for them.
Autonomous decisioning isn't deployable if it breaks compliance. Canon enforces RG and jurisdictional rules as part of every decision — and produces a regulator-ready audit trail by default.
- Self-exclusion list…
- Jurisdictional spend cap…
- Loss-chasing window…
- Stake escalation…
- Session duration…
Per-decision RG
Every decision passes through responsible-gaming checks.
Loss-chasing patterns. Stake escalation. Session duration. Deposit decline. Late-night first-time-deposit anomalies. Each signal is evaluated before a decision ships — at-risk players get bonuses held or blocked, never nudged harder.
- Loss-chasing detection (rolling-window stake / loss ratio)
- Stake escalation detection (per-session and rolling)
- Session duration limits with cool-down
- Deposit-decline pattern flagging
- Late-night and rapid-deposit anomaly detection
- Operator-set self-exclusion lists, honoured atomically
Four pillars.
Every agent decision passes through responsible-gaming guardrails. At-risk players get bonuses held or blocked, not nudged harder.
Every decision is logged with the signals, model version, and policy that produced it. Regulators can trace exactly why a player got what — at a per-decision level.
Bonus rules, wagering requirements, age and identity checks, and spend caps adapt to each player's market. New rules ship as policy updates, not engineering work.
Set spending caps, frequency limits, cool-down windows, and exclusion lists. Canon optimises within them — never around them.
FAQ
Safety, in detail.
UKGC, MGA, Spelinspektionen (Sweden), AGCO (Ontario), MGCB (Michigan), and most major EU regulators. New jurisdictional rule packs ship as policy updates.
Want a compliance-team walkthrough?
We do RG-officer-led demos that focus on guardrails, audit, and jurisdictional rules — not retention lift.