AI Assistants Failing: Research Shows Chatbots Directing Users to Unlicensed Casinos

Artificial Intelligence is frequently hailed as the ultimate solution for online safety, yet in the realm of online gambling, it appears to be doing the exact opposite. A recent study from the Netherlands has uncovered that popular generative AI chat and search assistants are systematically steering users toward illegal operators. Most alarmingly, these tools do not just name the casinos but often imply that these platforms are state-sanctioned and officially licensed. The risk to consumers is substantial, as critical protection mechanisms such as the Cruks self-exclusion register are completely bypassed by these AI recommendations.

The investigation highlights a dangerous flaw in the programming of systems like ChatGPT, Google Gemini, and Microsoft Copilot. Instead of bolstering the legal market, AI tools often emphasize the benefits of the black market, such as higher bonuses or the absence of identity checks. For players seeking a safe environment, searching with AI assistance has become a digital minefield. The legal gray area inhabited by many international operators is further blurred by these inaccurate AI responses, making the task of regulatory bodies significantly more difficult.

Numbers and facts

Researchers at XY Legal Solutions conducted their tests between August and September 2026. A total of ten leading AI systems were evaluated, including OpenAI's ChatGPT, Microsoft Copilot, Google Gemini, Anthropic Claude, and xAI Grok. The results are sobering: in 33 out of 79 recorded interactions, the tools named at least one operator lacking a valid Dutch license. The speed of the misdirection was particularly notable. Four programs—Mistral Le Chat, Google AI Mode, DeepSeek, and Grok—generated names of illegal websites after just a single prompt.

The experts flagged the presentation by Microsoft Copilot as especially problematic. The system generated a Top 50 legal casinos list that included numerous brands which, as of 1 September 2026, held no authorization from the Kansspelautoriteit (KSA). Despite initial ethical warnings from some AIs, many yielded results after repeated questioning. The study demonstrates that safety filters can be easily bypassed by reframing questions, effectively neutralizing consumer protections.

Background

The conflict over the visibility of illegal offers is not limited to AI chatbots. Parallel to the study, there is massive legal pressure on major tech corporations. The Dutch gambling trade association VNLOK has officially summoned Meta to court. The allegation is severe: the social media giant allegedly allowed widespread advertisements for illegal gambling to reach consumers on Facebook and Instagram despite repeated warnings. Legal proceedings were first initiated in June, and the situation has now escalated to the Dutch courts.

"We have repeatedly pointed out the scale of the problem to Meta and asked them to cooperate structurally in finding a solution. That has yielded insufficient results. Therefore, we are now submitting the case to the Dutch courts." - Björn Fuchs, Chairman of VNLOK

This development indicates that the industry is at a crossroads. While authorities attempt to control the market through licenses and exclusion registers like Cruks, tech platforms are opening new backdoors for unregulated providers through algorithms and AI errors.

Why it matters for German players

For players in Germany, this issue is highly relevant even though the study primarily focused on the Dutch market. In Germany, the State Treaty on Gambling 2021 (GlüStV 2021) strictly regulates who is allowed to offer online betting and slots. Anyone asking an AI for the best casinos risks being directed to providers with MGA or Curacao licenses that do not have permission for virtual slot games in Germany. The consequences are severe: these providers do not integrate with the cross-provider LUGAS tracking system or the 1,000 Euro monthly deposit limit.

Furthermore, the 1 Euro per spin limit on German online slots is a legal requirement for player protection, which illegal providers often advertise as a competitive advantage (no limits). German users should never rely on AI-generated lists and instead exclusively consult the official whitelist of the Gemeinsame Glücksspielbehörde der Länder (GGL). Only then can they ensure that strict German standards are met and player funds are legally protected.

What it means for GGL-licensed casinos

For legally operating casinos with a GGL license, AI misdirection represents an unfair competitive disadvantage. While licensed operators adhere to expensive regulatory requirements, they are often displaced in AI rankings by illegal competitors who can lure players with higher bonuses. The study proves that AI models like Mistral Le Chat often promote illegal sites by highlighting product benefits such as rapid payouts without identity checks. This undermines the GGL's efforts to channel the gambling urge into regulated and monitored paths. It remains to be seen whether regulators will eventually have to issue guidelines for AI developers to prevent the dissemination of illegal gambling content.