2026 will be the year when AI becomes a regulatory challenge, after the last 18 months have seen governments move from principles to enforcement. For instance, the EU AI Act — the world’s first comprehensive AI law, which begins its phased application on August 2, 2026 — introduces a framework where ‘high-risk’ systems face documentation, human oversight, and traceability obligations to mitigate harm and ensure accountability. This means training a model on the scale of GPT-4o now places a company within a compliance regime as demanding as the General Data Protection Regulation.

‍

Beyond Europe, lawmakers are also taking action. South Korea’s AI Basic Act, coming into force in 2026, introduces a “trust-by-design” certification, while in the U.S., state-level AI bills are emerging in states such as Colorado, Texas, and California, each defining developer duties and consumer-disclosure rights. Meanwhile, China has tightened real-time content and watermarking rules for generative models. For any company deploying AI across different jurisdictions, including those in the iGaming industry, compliance is becoming a geographical challenge, not just a technical one.

‍

The ethical dimension of AI is no longer theoretical, since regulators now expect decision traceability: the ability to show how an algorithm reached an outcome. In the iGaming industry, for instance, an AI-driven risk-management system that flags a user for exclusion must produce a verifiable audit trail: data input, model version, rationale, and human-review logs. Bias in automated systems is also moving from an ethical concern to a compliance issue. Under emerging fairness regulations, even unintentional patterns in AI-driven bonus allocation or player segmentation could be seen as discriminatory.

‍

Another emerging concern is liability. The EU had proposed an AI Liability Directive to extend civil liability to harms caused by autonomous systems, but the initiative was withdrawn in 2025 after lawmakers failed to reach agreement. Its withdrawal leaves a gap: companies remain responsible for proving that their AI systems act fairly and transparently, yet there is still no unified EU framework defining how liability should be shared between developers and deployers. For sectors like sports betting, this means legal responsibility for AI-driven decisions — from player-risk models to automated content — continues to rest on national laws and regulator interpretation.

‍

At the same time, open-source AI — designed to democratize innovation — could become the next compliance blind spot. Under the EU’s General-Purpose AI Code of Practice,providers of general-purpose models are expected to document data provenance and embed risk-management processes. iGaming companies that fine-tune such models without proper governance will find that “open” no longer means “exempt.” While the Code is voluntary, overlooking these practices could significantly increase compliance risk and negatively impact companies.

‍

Looking ahead, three trends stand out for 2026:

• Mandatory incident reporting. Under the EU AI Act, providers of high-risk systems will be required to report serious AI incidents to national authorities, much likedata-breach notifications
• AI certification. “Safe-AI” seals and independent audits will emerge
• Ethics as differentiation. Firms that embed transparency and accountability earlywill find compliance becomes a trust advantage.

‍

The coming year will test whether companies truly understand that ethics, regulations, and performance are now intertwined. In AI, responsibility will become the new innovation frontier for those who want to lead the iGaming market.