Every AI system deployed on the GAMB·8 platform is listed below with its risk classification, purpose, training data, and oversight mechanisms.
GAMB·8 Coach Overlay
Limited Risk
Purpose
Real-time gambling advisory surfaced as a transparent overlay during active sessions.
Training Data
Fine-tuned language model on a curated corpus of gambling Q&A pairs.
Human Oversight
User can dismiss any recommendation at any time. Training mode hides advice until after user acts.
Edge Radar
Minimal Risk
Purpose
Market edge detection across sports betting and casino verticals.
Training Data
Rule-based pipeline combined with a sports ML model trained on historical odds data.
Human Oversight
All outputs are advisory. User controls whether to act on detected edges.
Tilt Monitor
Limited Risk
Purpose
Behavioral anomaly detection to identify harmful gambling patterns and prevent harm.
Training Data
Behavioural rule engine (bet frequency, loss chasing, session length, and related signals) supplemented by a lightweight ML classifier.
Human Oversight
Users can dismiss any warning. Alerts are informational only — no automated account actions.
Multi-Agent Debate
Limited Risk
Purpose
Bull / Bear / Judge structured debate to analyse betting opportunities from multiple perspectives.
Training Data
Prompt-engineered templates running against commercial LLM APIs. No fine-tuning; fully prompt-driven.
Human Oversight
Advisory only. Full debate transcript is always visible. User makes final decision.
Poker Coach
Minimal Risk
Purpose
GTO strategy guidance across 13 analytical tools including range grids, ICM, and equity.
Training Data
CFR solver-validated range tables; deterministic analytical tools with no learned weights.
Human Oversight
All suggestions advisory. User can view range source and confidence for every recommendation.
Plan Checker
Minimal Risk
Purpose
Monte Carlo gambling plan simulation to assess bust probability and expected value.
Training Data
Deterministic Monte Carlo model. No ML model involved.
Human Oversight
User controls all inputs. Results include full methodology and confidence intervals.
Compliance Principles
The four pillars that govern how every AI feature on the platform is designed, deployed, and monitored.
Transparency
All AI recommendations display their confidence level and reasoning methodology. Users always know when they are interacting with an AI system — we never obscure this fact.
Human Oversight
Every AI feature includes explicit dismiss and override controls. No automated decisions are taken without user confirmation. Humans remain in the loop at every step.
Data Minimization
Behavioural signals are processed on-device wherever technically feasible. Server-side analytics events are retained for 90 days. We collect only what is strictly necessary.
Harm Prevention
Tilt detection, session limits, loss alerts, and responsible gambling tools are first-class features built into the core platform, not optional add-ons.
Your Rights
Under the EU AI Act and GDPR, you have the following rights with respect to how AI systems interact with your data and decisions.
Right to explanation — Click any AI recommendation to see the full reasoning behind it, including the model tier, confidence level, and data sources used.
Right to opt out — All AI features can be individually toggled off via Settings → Privacy. Opting out does not affect core platform functionality.
Right to erasure — Request permanent deletion of all your data, including AI interaction history and behavioural signals, at Settings → Account.
Right to human review — If you believe an AI system on GAMB·8 has produced an incorrect or harmful output, email support@gamb8.com to request manual review by a human agent.
Right to not be profiled — Disable all behavioral tracking (used by the Tilt Monitor and Edge Radar) in Settings → Privacy. Disabling tracking switches these systems to fully rule-based operation.
Legal Basis
Our legal framework for processing personal data and deploying AI systems within the European Union.