Image: © ronstik /Stock.adobe.com
The team at William Fry explain how game developers can utilise AI while keeping in compliance with regulations.
Ireland’s video game industry is growing rapidly. The Gaming in Ireland 2024 report by advertising agency TBWA revealed that the industry generated €465m in 2023, with revenues poised to reach €813m by 2027.
This growth is driven by a diverse gamer demographic in Ireland, with more than half of all adults in Ireland playing video games – surprisingly, 66pc of adult gamers are over the age of 35.
In 2022, the Irish Government launched a Digital Gaming Credit, giving companies a tax credit equating to 32pc of their game-creation expenditure. IDA Ireland says that the tax credit’s implementation “coincides with a shift in the way games are developed”.
“Unlike finite movie productions, games are now geared towards ongoing development over a long term, often involving hundreds on a team to produce engaging, immersive experiences that keep gamers coming back for more.”
While credits and Government support make game creation in Ireland more lucrative, the rapid development of artificial intelligence offers gaming companies another avenue to speed up game creation, all while raising several legal and ethical considerations along the way.
The many pros and cons
Laura Casey, an associate in the technology department at William Fry, explains that AI technologies enhance game design through “procedural content generation”, by creating “vast and unique game worlds and storylines with minimal input”.
AI also plays a “crucial role in improving gameplay” by developing intelligent NPCs, or non-player characters, Casey adds. In a video game, NPC actions are coded through software to react to a player’s action.
Traditionally, this relies on a pre-established script written by game developers and acted by voice actors. Now, a voice-to-text generative AI model can be used to expand the scope of what NPCs can say and do – creating a more immersive environment for gamers.
Moreover, the technology can help in other background tasks in game development, Casey says, including streamlining game testing and quality assurance, identifying bugs and performance issues more efficiently than human testers.
“The efficiencies introduced by AI have made the video games industry more inclusive, benefiting smaller games development studios by reducing costs and enabling them to create diverse and accessible experiences.”
Jennifer Hale, one of the most popular video game voice actors in the industry, explained in an interview with SiliconRepublic.com earlier this year that in her line of work, AI models are used mostly in the scratch process – the iterative process before actors are called in for their performance.
For voice work specifically, companies often need to hear how a script sounds before they begin recording the final product and that’s where most of the cost cutting using AI happens.
Unsurprisingly, AI in game development gives rise to a host of intellectual property (IP) rights issues, “especially around infringement and ownership,” according to Casey.
“If a games company’s AI outputs resemble protected works, it could face infringement claims, potentially impacting both the AI developer and the publisher. This isn’t just about copyright works – trademarks, patents and trade secrets are also in the mix.”
Massive developments in the technology in recent years mean that companies can now use AI to eliminate human costs by artificially synthesising voices and likeness by modifying already what already exists.
‘Legal maze’
Casey calls the ownership of AI-generated works a “legal maze”.
“In the US Copyright Office and US courts for example, it has been determined that AI-created works cannot obtain copyright registration.”
However, these ownership issues haven’t been tested in Irish courts yet, she explains. So, Irish law “might allow IP rights for AI outputs, defining the author as the person who made the necessary arrangements for the work’s creation”.
Casey advises gaming companies to “ensure AI-generated works have enough human input to secure IP protection across different jurisdictions”, in order to mitigate any IP-related legal issues.
“They also need to consider potential barriers like third-party IP rights, licensing terms and any restrictions on IP ownership of AI outputs before deploying AI.”
Rachel Hayes, a William Fry partner in technology, adds: “Intellectual property considerations are also crucial when using AI in gaming, particularly regarding the ownership and copyright of AI-generated content like characters and storylines.”
Hayes explains that AI models can monitor player behaviour in-game and will likely be used for age-assurance checks, which makes it important for video game companies in the EU to adhere to the region’s laws like the Digital Services Act and the GDPR.
“To align with the GDPR, the European Data Protection Board’s (EDPB) opinion suggests that companies developing or deploying AI models should consider implementing data minimisation,” she says.
This would include limiting personal data collection only to what is required and ensuring that data is anonymised to protect player identities and ensure the privacy of gamers.
“Transparency measures are also essential, providing clear and accessible explanations of how personal data is used in AI systems.”
The EDPB opinion, Hayes says, “is a reminder that innovation in AI must not come at the expense of players’ privacy rights”.
Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.
