Anthropic restricts Claude from creating clones of existing games

The rise of AI-generated 'clones' has become a growing concern, with users increasingly attempting to use generative technology to produce remakes of established titles. To mitigate the risk of infringing on copyrights, Anthropic has added specific instructions to the Claude Code prompt on the Opus 5.5 model. These updates are designed to steer the AI away from recreating specific, existing games, as such actions violate the company's usage policy.

This restriction does not mean a total ban on game development assistance. According to Anthropic staffer Addy Osmani, the tool will continue to support the creation of original games, genre-based content, and projects built from a user's own designs and assets. The company stated it is working to find the right balance between utility and legal compliance.

The decision follows increasing pressure from the gaming industry. Major studios, such as Bethesda, have previously had to take legal action to remove AI-generated versions of their properties, like Fallout, from the internet. While Anthropic is taking a proactive stance against direct cloning, the broader debate regarding how AI models are trained on scraped web data remains a complex and unresolved issue for the industry.

Despite these safeguards, the move may not fully alleviate the concerns of original game developers. There remains a risk that bad actors could use AI to rapidly produce low-quality 'slop' versions of game demos, making it difficult to fully protect intellectual property in the fast-paced era of generative AI.