AI, intellectual property & game development — key insights from Games Tech Connect

As we move into 2026 and look back at successes in 2025 for the games and interactive entertainment team, we report on Games Tech Connect — a brilliant event in Liverpool in October 2025 that we had the opportunity to be part of. It was hosted by FORMAT and the Liverpool Game Dev Network (LGDN) and brought together business leaders, developers, innovators and service providers working in the games and creative-tech sectors. 

Partner and Chartered Trade Mark Attorney Sara Ludlam and Associate Andreas Petrou from our commercial and IP team joined the speaker line‑up to talk about the use of generative artificial intelligence (AI) in video game development. Alongside the creative opportunities, we explored the tricky legal and practical challenges this raises — particularly around intellectual property (IP) ownership and how AI is reshaping the risks that developers face.

Here, Sara, Andreas, Paddy Fearnon and Lauren Hicks outline the key points from their session.

IP in game development — an overview

To start, we looked at what IP means in the context of game development and how it underpins the many different types of creative work involved — from software code and storylines to music, artwork and character designs. 

Types of IP

In the UK, IP is protected by several laws:

• Patent Act 1977.
• Copyright, Designs and Patents Act 1988.
• Trade Marks Act 1994.
• Registered Designs Act 1949.

Patents aren’t generally available for software in game development due to the difficulty of meeting the ‘technical effect’ requirement. 

That said, major studios like Nintendo and EA have secured patents for specific innovations. Nintendo has patented game mechanics like summoning and auto-battling systems (albeit these are currently being re-examined by the US Patent and Trademark Office) while EA has focused on technical and accessibility features including speech recognition and photosensitivity analysis. These moves have drawn pushback in the industry, particularly from smaller developers. 

If a studio’s software might qualify for patent protection, it’s essential to keep it confidential before a patent application is filed — so no pilot testing. We can provide further assistance on patentability and the preparation of confidentiality agreements if required. 

Where software doesn’t produce a technical effect, studios must rely on copyright to protect original code. Copyright is the most common IP right when it comes to IP protection for video games.

Copyright protects the expression of ideas — not the ideas themselves. For game developers, this typically includes content such as:

• Software/code.
• Storylines and in-game text.
• Music and sounds.
• Artwork and character designs.
• Films and gameplay videos.

Who owns the IP?

Ownership of IP in game development depends largely on the circumstances in which it’s created. If you produce IP in the course of your work for an employer, it’ll be owned by your employer (unless your contract says otherwise). However, if IP is produced outside of the scope of employment, it’ll usually belong to the individual creator — provided that they’re a ‘qualifying individual’ and the work meets the relevant legal requirements.

IP can be transferred but involves a formal process. Any transfer must be set out in writing, signed by the creator or previous owner and supported by consideration — for example, a lump-sum payment. 

Generative AI in game development

We then turned our focus to how generative AI is reshaping game development, looking at who owns AI‑generated output and the risks it creates for studios.

Who owns AI-generated output?

Most AI platforms assign output ownership to the user but refuse responsibility for infringement.

For example, OpenAI’s Europe Terms of Use (Updated 29 April 2025) contain the following statements: 

“…you are prohibited from:….Using our Services in a way that infringes, misappropriates or violates anyone’s rights; 

“You are responsible for Content, including ensuring that it does not violate any applicable law or these Terms. You represent and warrant that you have all rights, licences, and permissions needed to provide Input to our Services.”

“Ownership of content. As between you and OpenAI, and to the extent permitted by applicable law, you (a) retain your ownership rights in Input and (b) own the Output. We hereby assign to you all our right, title, and interest, if any, in and to Output.” 

In short, you own the output but you bear all of the risk — including if someone else’s output incorporates your input.

Consent is therefore critical. If AI-generated content is based on third-party works, you must have a written licence from the original rights holder. 

Always ask: 

• Do I have consent to input this material into a generative AI system?
• If I own the IP, do I want the system to use my input for other users?

Without ownership or consent, inputting material into a generative AI system risks infringement and potential legal action, though it’s important to note that different generative AI models may approach these issues in different ways. 

We also discussed another example — the use of generative AI for producing video content. OpenAI’s video tool ‘Sora 2’ initially launched with an opt-out model for IP owners. This meant that it could use IP protected works unless the IP owners had opted out. Users quickly began generating videos featuring well-known copyrighted characters, raising serious concerns about infringement. It’s been reported that in response, OpenAI shifted to an opt-in model for rights holders i.e. requiring rights owners to positively opt in to permit to their works being used. Even with these safeguards, however, there’s no guarantee that outputs won’t infringe someone else’s IP.

Recently, the Content Overseas Distribution Association (CODA), a body which represents leading Japanese game developers, including Bandai Namco and Square Enix, called on OpenAI to stop using members’ content in Sora 2, arguing that it infringes their IP rights. How OpenAI will respond to these demands — which are reflected by Japan’s government — remains unclear.

What are the risks?

1. IP infringement

One of the biggest risks with AI-generated works is the difficulty of proving originality. This creates a heightened change of being subject to third-party IP infringement and resulting compensation claims. Without clear evidence of how content is created, studios may struggle to defend against infringement claims or establish ownership rights. The financial consequences of IP infringement — whether actual or alleged — could be severe, with even legal costs of defending an alleged IP infringement claim potentially being enough to sink a small studio.

To protect your rights and defend effectively against claims, keep detailed audit trails showing your creative process — including inputs and outputs. Always document your workflows, check licences and ensure that you can explain how your content was created or, if using another person’s IP, to show that you have the rights to do so.

2. AI hallucinations & ‘making up’ information

A critical weakness of AI systems is that they don’t indicate when they lack knowledge. Instead of stating “I don’t know”, they can generate plausible but false information known as ‘hallucinations’. This is a major risk for developers relying on AI for code, assets or narrative content.  This makes it all the more important to always verify outputs and never assume AI-generated material is accurate or original.

Real-world consequences

Although not concerning the use of generative AI, a notable recent example of alleged copyright infringement in game development is Juracek v Capcom Co., Ltd. In this case, photographer Judy Juracek sued Capcom for using dozens of her original images that were published in Juracek’s book ‘Surfaces’ without consent. The images were used in games like Resident Evil. 

The case highlighted how unauthorised use of third-party content can result in multi-million-dollar settlements, legal costs and reputational harm. It serves as a clear reminder for game developers to secure proper licences and maintain robust audit trails for all creative assets. 

Contractual considerations (including investment & exit)

We then shifted the discussion to the contractual side of game development and how the use of generative AI may be concerning in light of typical contractual requirements.

Development & publishing agreements

Development and publishing agreements in the games industry drafted by software publishers will nearly always include warranties and indemnities relating to IP. A warranty is a contractual promise or assurance made by one party to another regarding the existence of certain facts or circumstances. If breached, it may result in the publisher being entitled to claim compensation from a development studio and — depending on the provisions of the contract — may entitle the publisher to terminate. 

Software publishers typically demand two key warranties, namely that:

1. the studio owns all relevant IP, and
2. the IP doesn’t infringe the rights of any third-party. 

Studios should try to resist the second warranty as it’s difficult to establish whether any third-party IP rights have been infringed. That said, this may be a commercial risk that a studio has to accept if the publisher is unwilling to negotiate. This risk may be even greater where generative AI has been used, particularly where the relevant input can’t be verified as non-infringing. These claims can be substantial and — in some cases — threaten the financial stability of the studio itself. 

In publisher-drafted agreements, warranties are often backed by indemnities — additional obligations requiring the studio to compensate the publisher if losses arise (such as from infringement claims). Indemnities increase the financial risk on the studio and should be resisted or negotiated where possible, though publishers may be reluctant. 

More recently, we’ve seen some publishers include an obligation that prohibits the use of generative AI entirely in the production of works, aiming to remove the risk of AI-related infringement as much as possible. This approach may become more common across the industry moving forward. As a result, studios must decide whether to accept these obligations (and the risk of breach) as drafted or push to negotiate exceptions that better align with the development process and the potential use of generative AI tools.

Investment & exit

Similar warranties and indemnities will also usually be included in the legal documentation relating to any investment into the studio or in any agreement for sale. Investors or buyers will usually carry out due diligence before completion of the deal, assessing and investigating the business. In higher value deals (though not limited to them), this will often include detailed technical due diligence.

Where technical due diligence is carried out, any issues around IP ownership (for example, linked to the use of generative AI) are more likely to be identified at that stage. This can happen through technical reviews of software products and any relevant tech stack or discussions with development teams. If such issues arise, they can cause delays, renegotiation (such as price reductions) or — in the worst-case scenario — abortive deals. 

If no issues are discovered prior to completion of the deal, the risk of breach of warranty and potential compensation claims against the studio will remain. 

Managing the risks

Internal training

• Mandatory training — provide compulsory generative AI training for all employees and personnel.
• Copyright awareness — ensure that staff understand that AI‑generated content may be subject to existing copyright.
• Communicate the risks — infringement, impact on funding, disruption to development deals and potential threats to the studio’s solvency.
• Tool‑specific guidance — set clear rules on which generative AI tools may be used, how they can be used and in what contexts.

Working with investors, buyers & partners

• Early declaration — disclose any use of generative AI at the outset and carry out a risk assessment to identify potential issues.
• Third-party content — identify all external content used and ensure that appropriate IP assignments or licences are in place. Always disclose any third-party material not owned or licensed by the studio.
• Audit trails — be prepared to explain the creative process. Keep detailed records of inputs, prompts and outputs and implement systems to track generative AI use for retrospective analysis if needed.
• Deal considerations — assess the risks of using generative AI against the potential impact on current or future deals, particularly where IP warranties or outright prohibitions on AI use may apply in agreements.

Talk to us

We support founders and studios with managing IP, AI risks, technology contracts and any other legal aspects of game development.

Our offering includes: 

• Advice on development, publishing and licensing contracts.
• IP portfolio assessment, registration and monitoring.
• Company setup and shareholder agreements.
• Investment funding and pitch support.
• Employment contracts and policies.
• Regulatory advice on (for example, regarding data protection and online safety).
• Dispute and reputation management.
• Company and asset sales.

Talk to us by giving us a call on 0333 004 4488, sending us an email at hello@brabners.com or completing our contact form below.