AI in gambling: Promises vs. reality

Moody’s Outlook 2026 highlights a critical gap: artificial intelligence in iGaming consumes significantly more resources than it generates in profit. Operators invest in experience personalization, behavioral analytics, security, and marketing, yet only 5% of projects achieve financial returns. The remaining projects burn budgets on computing power, data engineering, and opaque algorithms that are difficult to interpret or improve.

AI effectiveness is limited to specific tasks. Martins Lielbardis of the iGaming Centre notes success in risk scoring, support automation, product testing, and anti-fraud. However, without high-quality data, long-term model training, and integration into operational processes, implementation turns into inefficient spending with no tangible business impact.

Algorithms are unable to replicate the emotional core of gambling—unpredictability, intuition, and psychological impulses. The UK Gambling Commission demonstrates the right balance: AI flags risks, overheated accounts, or anomalous activity, but decisions remain with humans, preserving the authenticity of the game.

Four areas of AI development in gambling require further elaboration with practical examples:

1. Experience personalization — systems adapt interfaces and offers to an individual player’s style. For example, platforms like Bet365 use AI for dynamic slot recommendations: if a player prefers high-volatility games with bonus rounds, the algorithm suggests precisely such titles, increasing retention by 15–20%. At the same time, the system tracks session length and suggests breaks when signs of fatigue appear.

   
   

2. Behavioral analysis — risk monitoring for responsible gambling. Kindred Group (Unibet) has implemented algorithms that scan 200+ parameters: betting frequency, loss size, time of day, and pattern changes. When “red flags” are detected—for example, 10 deposits in a day or play after 3:00 a.m.—the system sends personalized reminders and temporary limits, reducing complaints related to gambling addiction by 27%.

   
   

3. Security and anti-fraud — combating illegal activity and bots. Evolution Gaming uses AI to analyze network patterns: algorithms detect multi-accounting (one IP, multiple devices), wash trading (artificial turnover), and AI bots that play with statistically perfect behavior. In 2024, this blocked fraudulent networks worth €45 million, preserving the platform’s reputation and licenses.

   
   

4. AI marketing — automated content without manipulation. LeoVegas applies generative AI for personalized email campaigns: instead of mass mailings, players receive messages like “Your favorite slots with 20% cashback,” delivered at precisely the right time. Conversion rates increase by 32%, while the algorithm complies with UKGC limits—no more than three contacts per week.

   
   

Risks outweigh potential. AI costs are growing exponentially, while monetization lags behind. Operators become dependent on vendors from Malta or the UK, surrendering control over data. “Black box” systems complicate auditing.

Only a standardized approach with ethical oversight and interdisciplinary teams can transform the technology from a cost center into a catalyst for sustainable growth. Gambling remains an exclusively human phenomenon that no algorithm can replace.