Major chatbots linked to unlicensed gambling sites
Table of contents
- A two-week investigation by Investigate Europe, from March 9, found that major AI chatbots recommended unlicensed gambling sites in roughly three-quarters of responses across 10 European countries.
- Meta AI, Google Gemini, and ChatGPT led the unlicensed casino recommendations, with some chatbots also advising users on how to bypass self-exclusion schemes such as Gamstop and Germany’s Oasis.
- The findings arrive as Yield Sec’s 2024 report (published August 2025) shows that unlicensed operators already account for 71% of Europe’s online gambling market.
Prominent AI chatbots, including Meta AI, Google’s Gemini, ChatGPT, Grok, Microsoft Copilot, and Claude, are routinely directing users towards unregulated online casinos and providing guidance on circumventing consumer protection safeguards, according to a cross-border investigation published March 9, 2026 by Investigate Europe and media partners including The Guardian.
Over two weeks, researchers tested seven leading AI chatbots across 10 European countries, using 30 prompts per platform in national languages.
Requests covered casinos with the biggest bonuses and platforms that do not require age verification. The chatbots recommended gambling sites not licensed in Europe in roughly three-quarters of replies.
Blacklisted sites promoted
Testing across 10 European countries found that Meta AI recommended an unlicensed casino in 27 of 30 prompts, Gemini in 26, and ChatGPT in 22. By country, France and Poland each received 18 recommendations for unlicensed sites out of 21 prompts, with the UK producing 14.
Among the sites recommended were dozens that Investigate Europe found had been blacklisted or fined by European regulators for lacking domestic licences. In Greece, ChatGPT promoted ten casinos described as having “remarkable features”, all of which appeared on the Greek regulator’s blacklist.
Microsoft’s Copilot similarly recommended unlicensed online casinos in Austria.
Chatbots also absorbed and repeated marketing language from operators registered in offshore territories such as Curaçao and Anjouan, jurisdictions known for limited regulatory oversight.
In Spain, Grok described Curaçao casinos as offering “very generous bonuses”. In Poland, Gemini stated that “choosing a casino without verification is a popular trend in 2026 among players who value privacy and instant payouts”.
Two experts pointed to distinct but related failings. Paulo Dimas, CEO of the Center for Responsible AI, a Portugal-based research organisation, attributed the pattern to poisoned training data, arguing that AI models learn the marketing language of illegal gambling by ingesting vast volumes of promotional content from unlicensed operators.
Dimas said:
“Illegal gambling operators flood the internet with billions of pages of promotional content to hijack search results. Because the models ingest this ‘grey market’ content at scale, they learn the marketing language of illegal gambling as a primary dialect.”
Tiemo Wölken, a German MEP who helped draft the EU’s Digital Services Act, said the results demonstrated “emerging risks associated with AI chatbots”, warning they risk becoming supercharged search engines without the safeguards of traditional online search.
Self-exclusion schemes undermined
The investigation found a particularly acute risk for vulnerable players seeking to limit their gambling activity. When prompted for casinos not covered by self-exclusion schemes, Grok, Meta AI and Gemini recommended unlicensed operators in all countries tested.
One Meta AI response described no-ID casinos as the “Holy Grail”, promoting platforms that circumvent identity verification. Google’s Gemini said crypto sites offered players “anonymity” and a “lack of rigid limits”.
Schemes such as Gamstop in the UK and Oasis in Germany are mandatory for locally regulated firms, designed to allow individuals to block access to all legal online gambling in a given country. When all seven chatbots were prompted for the top ten online casinos not on Gamstop, all recommended websites capable of skirting the scheme.
Will Prochaska, spokesperson for the UK Coalition to End Gambling Ads, said:
“Promoting and praising illegal casinos for their ability to circumvent regulations undermines the rule of law and puts people in danger. We urgently require legislation mandating Know Your Customer checks for tech platforms before promoting any gambling operator.”
The UK Gambling Commission said it is aware of the concerns and is treating them seriously, describing protection of consumers from unlicensed overseas operators as a priority. Wölken added that the EU’s AI Act and Digital Services Act could be more robustly applied to address the systemic risks posed by AI models.
Not all platforms performed equally poorly. Claude, developed by Anthropic, was the chatbot most likely to avoid naming unlicensed operators in its answers altogether.
OpenAI said ChatGPT is trained to refuse requests that facilitate illicit behaviour and that it had strengthened the model’s ability to identify emotional distress and direct users to real-world support.
Microsoft said Copilot uses multiple layers of protection, including automated safety systems and real-time prompt detection.
A market already under pressure
The investigation lands against an already challenging enforcement backdrop. According to Yield Sec’s 2024 report (published August 2025), commissioned by the European Casino Association, illegal gambling operators captured 71% of Europe’s online gambling market, generating €80.6 billion in gross gaming revenue — more than double the €33.6 billion produced by the licensed sector.
That represented a 53% year-on-year increase in illegal online GGR. At an assumed EU-wide online gaming tax rate of 25%, illegal operators diverted more than €20 billion in tax revenue away from European economies in that year alone.
Youth exposure adds another dimension. A 2024 study by GambleAware found 87% of UK children had encountered gambling content online, with unlicensed operators identified as a growing part of that exposure.
A European School Survey Project on Alcohol and Other Drugs found almost a quarter of 15- to 16-year-old students had gambled for money in the past year, with harmful gambling behaviours on the rise among teenagers.
For regulators and operators, the findings sharpen pressure on both AI developers and policymakers to close what Wölken called a regulatory gap before generative AI consolidates its role as a primary gateway for consumer decisions in the online gambling market.
Keep reading
Do you have a story worth sharing?
Send it over to our editors!
