Does AI have a Winning Hand at the Gambling Table? A General Look at AI implications in Gambling 5v1g1r
Cosmina Simion, Managing Partner WH Simion Partners 35734
By Cosmina Simion (Managing Partner) and Cătălin Veliscu (Junior Associate) WH Simion & Partners
Cătălin Veliscu (Junior Associate) WH Simion & Partners
The integration of AI in the gambling field was not a matter left to chance. In a field where technology is at the core and stakeholders are always looking for innovative solutions to address ongoing problems, the use of AI should come as no surprise. However, is the implementation of AI in gambling just another example of using AI just for the sake of it? Whether this is another instance of the hype bubble around AI[1] or a beneficial use in a very sensitive field, the implications are vast, and, in this article, we will try to tackle some of these issues from a generally legal perspective.
For this article, we did not limit our research to a specific branch of gambling. Rather, we tried to assess AI’s uses in as many cases as possible with the purpose of determining if its implications are advantageous or if it enhances the inherent risks of this industry. We used as a guiding principle the interaction of gambling operators with their players, based on AI’s unrivaled power of analyzing large data pools. From a purely business-oriented perspective, it may seem that automated solutions can only boost profits and generally eliminate human errors from business decisions. However, the great challenge comes when assessing the consequences of using AI in relation to the customers, more so in an industry which has a duty in properly protecting its clients. We focused on three specific issues and that is why our first part deals with responsible gambling and the role of AI in both aiding and neglecting this effort (Section 1). Moreover, the second part looks at current trends in gambling advertising which is more and more centered on a personal experience of the player with the brand (Section 2), while the third part takes a step back and analyzes whether AI can properly represent a solution for operators with respect to regulatory requirements (Section 3).
Section 1 – Protecting players and mitigating vulnerability risks 2nw6h
The main advantage of using AI, irrespective of field, is the power to analyze large data banks. Of course, legal implications can appear both at this stage and at the ulterior moment when the results are used for different aims. With respect to addressing the responsible gaming issue, using AI to assess individual player profiles, especially in online betting can help preemptively detect risky patterns in playing, which in turn can determine the operator to alert the player or direct him to resources meant for addressing addiction. Risky player patterns that should trigger a careful approach from the operator can be represented by reversing withdrawals, continuous betting/playing after losing streaks, impulsive behavior and spending etc.[2]
Instead of a reactive or post factum approach to these issues, ethical use of AI in gambling can lead to a proactive strategy for operators to deal with responsible gambling and excessive harm. It can help identify early signs of a vulnerable player based on their history or the market in which they play. In turn, operators can set up thresholds for these signs based on this data, can justify their decisions on a bigger data pool and can reduce the resources that they invest in monitoring these aspects.[3]
Besides the summarizing powers of AI, gambling can also benefit from generative models that use machine learning to generate content based on inputs. In fact, game development can be enhanced with generative AI (for coding, graphics, game concepts etc.) and marketing schemes (see Section 2). The challenge of this player-based approach is the limiting of addiction concerns, generated by a bigger retention on the operator’s app or website which, consequently, lead to a bigger risk of harmful gambling.
The use of generative AI can raise issues regarding the operator’s rights on the final product, as, at the moment of writing this article, no EU jurisdiction allows for purely AI-generated content to be protected under copyright. Moreover, intellectual property concerns can also be linked to the training data used by that specific model and whether the producer of the AI used the works lawfully. Some of these issues will be addressed with the entry into force of the EU AI Act, with the first set of rules taking effect from February 2, 2025. Briefly, under the AI Act, gambling operators who use AI systems will be qualified as deployers and will need to comply with the regime imposed by the Regulation.
AI
Section 2 – Personalized player engagement and predictive analytics 6d6140
Even though the whole advertising industry has shifted to a personalized approach when engaging with customers, advertising in gambling leads to legal challenges regarding addiction prevention and responsible gambling in general. Even more so, using the processing power of AI system for this purpose must be done ethically and in line with the existing legal framework, even though AI may not be directly addressed. For example, in sports betting, AI can be used for analyzing historical and current data like statistics or weather to predict game outcomes[4] or to determine odds in a more facile manner. Operators can use now systems based on machine learning to suggest different bets to players, based on their history and their risky/cautious approach.
This approach is concretely implemented through tailor-made offers for players, like bonuses or VIP schemes made to appeal to certain players. Also, online platforms can use AI solutions to also analyze their relationships to more easily determine which agreements lead to more traffic on their platform[5]. Consequently, marketing schemes can be adapted to the interests of players attracted by specific agreements.
While maybe more subtle, even land-based operators can ensure lengthier player retention (the drawback being the responsible gambling concerns) through analyzing the behavior of past players. In this case, focus changes from the tendencies of the player to the interaction in itself with a specific slot machine or game.[7]
However, the main risk in regard to individual advertising is related to setting up data banks with information about players’ behavior. This may be problematic from a privacy perspective. To ensure compliance, players should be made aware that their gambling-related tendencies are being used for this purpose and operators should properly process this data, only for the declared purpose and through systems that ensure proper security.
WH Simion & Partners
Section 3 – AI standards applied to fraud prevention and general compliance 4lu5
AI can also be a for gambling operators in complying with legal requirements for fraud prevention and for KYC obligations. Based on the same pattern previously mentioned, AI analysis of player history and their tendencies can be used to detect anomalies and then flag these irregularities that can be proof of fraudulent behavior.
A similar preventive strategy used for marketing purposes can be used to determine these deceitful intentions. For instance, suspension or removal can be done faster, the human error risk is limited, and monitoring is autonomous and operated non-stop. Furthermore, AI has a big impact in reducing false positive results in either cheating cases or in money-laundering attempts, by optimizing at first the KYC procedure and then the monitoring of player behavior. Evidently, even though the employment perspective is not the object of this analysis, as in many other industries, the use of AI in gambling represents a risk for layoffs in the departments that are more prone to automatization.
While privacy and personal data concerns are inherent to any use of AI, the biggest risk from a fraud prevention perspective is in the need for human intervention and control. Although the technology is based on allegedly precise algorithms, false results or hallucinations[9] and it essentially means that humans cannot properly trace the system’s thought process, the ‘why’ behind its decision. Therefore, human intervention done by operators is of the utmost importance generally speaking, even more so in gambling.
[1] The AI hype bubble is the new crypto hype bubble, Cory Doctorow, Medium, March 9, 2023, https://doctorow.medium.com/the-ai-hype-bubble-is-the-new-crypto-hype-bubble-74e53028631e, accessed on November 21, 2024.
[2] Concern as the gambling industry embraces AI, Nicola K Smith, BBC, March 5, 2024, https://www.bbc.com/news/business-68304137, accessed on November 20, 2024.
[3] Applications of AI in the Gambling Industry, Dr. Simon Hembt, Toby Bond, Elizabeth Dunn, Callum Granger, Bird&Bird, June 13, 2024, https://www.twobirds.com/en/insights/2024/uk/applications-of-ai-in-the-gambling-industry, accessed on November 20, 2024.
[4] The Game Changer: How AI Is Transforming the World of Sports Gambling, Neil Sahota, Forbes, February 11 2024, https://www.forbes.com/sites/neilsahota/2024/02/11/the-game-changer-how-ai-is-transforming-the-world-of-sports-gambling/#:~:text=Moreover%2C%20AI%20plays%20a%20huge,resources%20to%20help%20maintain%20control, accessed on November 20, 2024.
[5] AI in gambling: how online casinos use artificial intelligence, Petr Stehlik, Slotegrator, June 27, 2024, https://slotegrator.pro/analytical_articles/how-ai-can-optimize-online-casinos/, accessed on November 20, 2024.
[6] Predictive AI And Slot Machines: Shaping The Future of Casino Gaming, Jared Thau, Forbes, April 26, 2024, https://www.forbes.com/councils/forbestechcouncil/2024/04/26/predictive-ai-and-slot-machines-shaping-the-future-of-casino-gaming/, accessed on November 20, 2024.
[7] Using AI/Machine Learning in Gambling Applications (Case Study of Machine Learning for Operational Decision Making in Casinos), Chima Omike, December 17, 2021, http://dx.doi.org/10.2139/ssrn.4025575.
[8] What are AI hallucinations?, IBM, https://www.ibm.com/topics/ai-hallucinations, accessed on November 21, 2024.
[9] AI’s mysterious ‘black box’ problem, explained, University of Michigan-Dearborn, March 6, 2023, https://umdearborn.edu/news/ais-mysterious-black-box-problem-explained, accessed on November 21, 2024.
