What artificial intelligence actually does in an online gambling business, how the models are built and governed, and what the law now requires of them. The course covers the distinctions the marketing blurs (statistical machine learning, generative models, relabelled automation) and the value families, customer value and personalisation (churn, lifetime value, propensity, next best action, recommendation, personalised promotions, experimentation, and the constraints from data protection, player protection and marketing rules), fraud, financial crime and integrity (bonus abuse and multi-accounting, payment fraud, account takeover, anti-money-laundering monitoring, sports integrity, sharp detection and stake factoring, and the adversary's own models), player-protection models (the contested target, features and streaming architecture, the interaction ladder and its evaluation, regulators' expectations, the single customer view, and the open questions), trading, content and operations (pricing models, generative customer service in tiers, content and localisation under the advertising codes, game design and testing, back-office uses, identity vendors, and an honest account of what works), and governance and regulation (GDPR automated-decision rules, the EU AI Act's phased obligations including the 2027 high-risk deferral, gambling regulators' expectations, the model-governance programme, and running generative tools responsibly).
Written for operators and suppliers deploying models, for compliance, legal and data-protection teams who must govern them, for product and data people entering the industry, and for anyone who wants to separate what AI does in gambling from what conference keynotes say it does.
The through-line is that the value is in the operator's own data and models, that the obligations attach to the operator's decisions rather than the supplier's software, and that protection must run ahead of commerce in every system that touches a customer.