The betting and iGaming industry, explained

Verte News

Edition: English

Business

ICE Research Institute shifts focus from funded studies to industry adoption

Director Ewa Bakun says the Barcelona-based institute wants its gambling research acted on, not just published, as its first practical tool reaches operators.

Ewa Bakun, director of the ICE Research Institute, pictured for iGaming Business
AI-generated illustration

The ICE Research Institute (IRI) is moving into a new phase of its work, shifting from commissioning academic studies into persuading operators, suppliers and regulators to put their findings into practice.

Ewa Bakun, the institute's director, told iGaming Business that much gambling research stalls once it reaches a journal or a conference stage. The IRI, she said, was set up to do something different: use its reach through ICE, iGB Live, iGaming Business and World Gaming Executive to put findings in front of stakeholders and work with them on adoption rather than simply funding studies and stopping at publication.

The institute grew out of ICE's relocation from London to Barcelona. Discussions between the show's organisers, the City of Barcelona and Fira Barcelona focused on what legacy the event could leave beyond its roughly €255 million annual contribution to the city and beyond the three days of the show itself. The result was an institute built around prevention, policy and technology.

It currently funds seven projects, spanning AI-based early detection of gambling-related vulnerability, adolescent education and media stigma. Every project is led by a Spanish institution, in several cases paired with an international partner such as the University of Nevada, Las Vegas. Bakun said the common thread is prevention rather than treatment, since by the time treatment is needed the harm has already occurred.

Some of the funded work asks difficult questions of the industry, including neuroscience research testing whether safer gambling messages actually change behaviour, and a study into why relatively few people use harm-minimisation tools at all. Bakun argued that any intervention should be judged on its measured effect rather than on the intention behind it.

The institute takes no gambling industry money and does not fundraise. Its ring-fenced budget comes through ICE's agreement with Fira Barcelona, while academic rigour sits with the universities carrying out the research. Bakun is now pausing new commissions to concentrate on completing current projects and is setting up two independent bodies to oversee future funding calls and grant selection.

The first project to produce something operators can use directly is the AI Use Case Registry, built by the UNLV International Gaming Institute's AI Research Hub alongside Universidad Rey Juan Carlos. The project followed the inaugural State of AI in Gaming report, published in April 2026, which found AI being deployed across back-office and customer-facing functions with very little governance in place.

Kasra Ghaharian, director of research at the International Gaming Institute, said the pattern echoed the early, unregulated expansion of social media and was something the sector should avoid repeating. He said the biggest gap was visibility, since companies cannot govern AI use they cannot first see across their own organisation.

The Registry is free, open source and available on GitHub. Organisations can adapt it to log every AI use case and document how each is governed, either departmentally or through a central oversight team. A screening step sorts each use into one of three tiers, with documentation requirements scaled to the potential impact on players.

The team reviewed existing frameworks, including the EU AI Act and the UK's Algorithmic Transparency Reporting Standard, and found both scaled oversight sensibly to risk but were pitched at a broad audience and hard to apply to a product with addictive potential. Ghaharian said many governance frameworks stop at high-level principles, whereas the Registry was designed for companies to implement immediately. Regulators, operators, academics and AI specialists tested and refined the tool through focus groups and two rounds of review before release.