Anthropic has agreed to maintain copyright protection measures in its AI systems as part of an agreement with music publishers. This development comes from their ongoing legal dispute in the Northern District of California.

The agreement requires Anthropic to:

• Maintain existing content filters on user queries
• Implement similar protections in future AI models and products
• Respond promptly to publisher notifications about potential copyright violations
• Investigate and address any reported copyright issues
• Provide detailed written responses about how issues will be resolved

While Anthropic can improve and optimize these guardrails, they cannot reduce their effectiveness at preventing copyright infringement. Publishers retain the right to notify Anthropic if they believe the protections aren't adequately preventing unauthorized use of their content.

This agreement partially resolves the preliminary injunction motion filed by eight music publishers in August 2024, following their October 2023 lawsuit. The publishers initially sought the injunction to prevent copyright infringement, while Anthropic maintained its use of copyrighted content for AI training qualified as "fair use" due to transformation of the original work.

The agreement explicitly states it should not be interpreted as an admission of liability or wrongdoing by any party. The broader issue of whether Anthropic can use unauthorized lyrics to train future AI models remains unresolved and pending in court.

Related Articles

Previous Articles