AI The game industry was an early and provide clarity for innovation and adopter of AI in its quest to provide will produce non-statutory guidance. enhanced gameplay and develop and While specific guidance will publish innovative games at speed undoubtedly be helpful and is already while reducing costs. Automation of being produced by some regulators, non-player characters (NPCs) and there are a number of issues to personalisation of gameplay have be aware of when using AI in or to long been AI-driven. As the uses develop games. of AI in games continue to evolve, developers and publishers must Copyright and ownership consider a variety of legal issues. AI models need to be trained. Where Unlike the EU whose new AI Act they are trained on existing creative looks to govern the use of AI top- works, questions of copyright will be down, the UK government decided engaged. This becomes particularly not to legislate to create a single relevant with generative AI models function to govern the regulation of which are often trained on datasets AI. It has instead elected to support and materials found online. existing regulators in developing a sector-focused, principles-based The UK government has confirmed it approach. Regulators including the will not be introducing an exemption Information Commissioner’s Office for text and data mining (equivalent (ICO), the Competition and Markets to the EU TDM exception). For now, Authority (CMA), the Financial Conduct AI developers training generative AI Authority (FCA), Ofcom, the Health models in the UK using unlicensed and Safety Executive and the Human third party materials would need Rights Commission will be required to to rely on the ‘temporary copies’ consider five principles to build trust exception. This applies to copies that are transient or incidental, an integral 154