By Tatiana Martins, journalist at G&M News.
The study you have launched, ‘The State of AI in Gaming 2026’, establishes the first global benchmark for AI adoption in the gambling industry. What were the main gaps in knowledge or perception that motivated the creation of this report?
Kasra Ghaharian: The core motivation was to move the conversation from anecdote to evidence. Industry didn’t have a reliable way to benchmark where they stood relative to peers. Regulators and other stakeholders didn’t have knowledge or data on how AI was actually being deployed. The report is designed to fill these and other gaps, providing a data-driven baseline to help guide the conversation on AI moving forward.
This document highlights a clear disconnect between ambition and execution, particularly with the AI Maturity Index averaging 45 out of 100. In your view, what are the biggest barriers preventing operators from scaling AI effectively?
Rick Arpin: There appears to be gravitation towards what the industry knows, versus what it can aspire to. This may be influenced by being in a highly regulated industry. In the land-based sector, this is also impacted by a complex systems environment, making it more difficult to harness data for AI applications. Viewed from multiple lenses, the trends are similar. For example, gaming companies are deploying simple AI applications compared to other industries (which are rapidly deploying agentic AI). In gaming, AI is being used more in technical and administrative areas than customer-facing areas. We believe industry needs to invest its efforts into changing the culture, experiment more rapidly, and, at the same time, establish governance programs. It’s not too late to catch up, but the industry can’t risk being left behind.
Governance emerged as the weakest dimension, with only a minority of companies having dedicated AI oversight. How critical is governance in this stage of AI adoption, and what risks does the industry face if this gap is not addressed quickly?
Kasra Ghaharian: I think the major risk here is trust. AI governance enables trust, whether it be from customers, regulators, or other stakeholders. Just as the industry maintains rigorous standards for AML and responsible gambling, we must now apply that same level of discipline and effort to AI as its development accelerates and its applications expand.
One of the most striking findings is the differences in criteria observed between regulators and operators regarding AI deployment. How can both sides work together to improve transparency, trust, and effective oversight in such a fast-evolving landscape?
Kasra Ghaharian: I believe there are several options here; many are likely not unfamiliar. We can look to established methods like sandboxing and collaborative working groups, but there is also a unique opportunity for “open” self-governance. At the UNLV IGI’s AI Research Hub (AiR Hub), we are tackling this with our AI Registry project. Building on research published last year by IGI Graduate Research Assistant, Cesar Lozoya Martinez, and with funding from the Clarion ICE Research Institute, we are currently engaging stakeholders to develop a framework where industry can proactively document AI use cases. This approach ensures transparency in high-risk areas and fosters a proactive rather than reactive relationship with oversight bodies.
While generative AI is already widely adopted, agentic AI remains limited due to the high-stakes nature of gambling operations. How do you see the role of more autonomous AI systems evolving within the industry, particularly in balancing innovation with responsible use?
Simo Dragicevic: Achieving the right balance between innovation and responsible use will be a challenge for the industry in some of the major use cases of agentic AI. Whilst there will be pressure to press ahead with deploying agentic AI in customer support functions to improve customer experience and to reduce costs, progress will be heavily dependent upon the industry demonstrating that AI can be trusted. Developing this trust will require evidence-based guardrails that do not exist today. Developing these will need collaboration between industry, regulators, academics, and various support sectors. Outside of this particular use, we observed developments in other parts of the industry, for example, in prediction markets. Here, agentic AI can make autonomous decisions, like selecting events, creating markets, and settling outcomes. In use cases like these, there is a clearer path to utilizing the technology without the barriers mentioned earlier and we expect to see continued innovation.