AI for Business

AI Chatbots Steer Users Toward Black-Market Betting Sites

A recent investigation has uncovered a serious flaw in popular AI assistants: they are regularly directing users to illegal gambling websites. According to a report from Digital Trends, when asked...

Share:

A recent investigation has uncovered a serious flaw in popular AI assistants: they are regularly directing users to illegal gambling websites. According to a report from Digital Trends, when asked questions like "where can I play online poker," both OpenAI's ChatGPT and Google's Gemini respond with lists that include unlicensed and explicitly blacklisted operators.

The issue is not sporadic but systematic. Researchers found the models recommending platforms that regulators in the UK, Australia, and Europe have flagged for operating without authorization. These sites often lack age verification, responsible gambling tools, and legal recourse for users - precisely the risks regulation aims to prevent.

Legally, this presents a significant problem. Promoting unlicensed gambling is a criminal offense in places like Britain under the Gambling Act 2005, with similar laws in the EU and Australia. An AI's recommendation could be interpreted as advertising, potentially implicating the tech companies behind them.

The root cause lies in how these models are built. They generate answers from patterns in their vast training data, which includes promotional internet content from unlicensed operators. Unlike Google's search ads, which require license verification, Gemini operates without such safeguards. OpenAI's own policies forbid generating content that facilitates illegal activity, yet the chatbot routinely violates this rule.

Legitimate gambling affiliates, who must perform rigorous license checks and geo-targeting, view this as a profound inequity. One consultant called it "the biggest unlicensed affiliate operation in history."

So far, responses from OpenAI and Google have been dismissive, framing it as an accuracy issue. But the sites recommended are real, not hallucinations. The core failure is the absence of any compliance mechanism. Experts argue the companies could implement jurisdiction-aware filters, as they do for other regulated content. Until they do, these widely used AI tools will continue to function as unaccountable referral engines for black-market gambling, creating real consumer harm while skirting the rules that bind everyone else in the industry.

Source: Webpronews

Ready to Modernize Your Business?

Get your AI automation roadmap in minutes, not months.

Analyze Your Workflows →