Legal Expert Warns Studios to Actively Protect IP as AI Integration Grows in Game Development

Outfit7’s legal VP warns that AI integration complicates intellectual property rights, making automatic ownership no longer guaranteed. She stresses the need for proactive documentation and human oversight to maintain control over creative assets. The discussion highlights risks around copyright, confidentiality, and the evolving nature of authorship in AI-assisted workflows.

On November 3, 2026, Tina Stockbauer Zagorac, VP of legal at Outfit7—the studio behind the globally recognized Talking Tom franchise—spoke out about the growing risks to intellectual property (IP) in game development due to generative AI integration. Her warnings center on how AI undermines long-standing assumptions about automatic copyright ownership, particularly when human input is minimal or untracked. She emphasized that, unlike traditional workflows where employee-created code or art automatically belonged to the employer, AI-assisted creation now demands active legal management to secure IP rights. This shift is especially critical for studios working with licensed brands or unreleased content, where feeding proprietary assets into public AI tools could result in trade secret exposure, breach of contract, or loss of confidentiality. Outfit7’s position reflects a broader industry concern as AI becomes embedded in design, coding, and marketing processes.

The background to this issue lies in the global legal principle that copyright requires human authorship—meaning AI-generated content without substantial human intervention may not qualify for protection under laws like the U.S. Copyright Act or EU directives. This creates a vulnerability: if studios fail to document human oversight, their assets could end up in the public domain or be legally unenforceable. Zagorac stressed that companies must implement strict protocols—such as disabling model training in AI tools, aud