Related Rights

Related rights, or neighbouring rights, are intellectual property rights distinct from copyright. While copyright protects authors of original works, related rights protect parties who play a crucial role in communicating those works to the public. These rights primarily cover three categories: performers (actors, musicians) for their performances; producers of phonograms (record labels) for their sound recordings; and broadcasting organizations for their radio and television programs. These rights are internationally recognized under treaties like the Rome Convention (1961) and the WIPO Performances and Phonograms Treaty (WPPT, 1996). For enterprise risk management, especially in AI, failing to clear related rights in training data (e.g., music, audiobooks, film clips) constitutes infringement, leading to significant legal and financial liabilities.

In enterprise AI risk management, managing related rights involves a structured process to ensure the legality of training data. Key steps include: 1) **Data Asset Auditing:** Systematically scan all datasets, particularly audio and video files, to identify content protected by related rights. Use metadata analysis to flag high-risk assets. 2) **Rights Clearance and Licensing:** For identified assets, determine the rightsholders (e.g., performers, record labels) and secure explicit licenses for machine learning and Text and Data Mining (TDM) purposes. This often involves negotiating with Collective Management Organizations (CMOs). 3) **Compliance Monitoring:** Implement automated content identification systems to continuously monitor data inflows and ensure no unlicensed protected content is used. A global tech firm developing a voice synthesis AI applied this process, achieving a 100% compliance rate for its audio training data and mitigating infringement risks.

Taiwan enterprises face several key challenges in managing related rights for AI: 1) **Complex Rights Ownership:** A single media file can have multiple rightsholders (performers, producers), making it difficult to trace and clear all necessary rights. **Solution:** Partner with established music licensing platforms or CMOs that offer bundled licenses. 2) **Legal Ambiguity for AI Training:** Traditional licenses do not explicitly cover Text and Data Mining (TDM) for AI, creating legal uncertainty. **Solution:** Engage legal experts to draft specific AI/ML clauses in licensing agreements and adopt a risk-based approach by securing the broadest rights possible. 3) **Global Data vs. Local Laws:** Using international datasets requires navigating a patchwork of different national laws on related rights. **Solution:** Establish a global compliance policy based on a high standard, such as the EU's DSM Directive, and classify data sources by jurisdiction-specific risk levels.

Winners Consulting specializes in related rights for Taiwan enterprises, delivering compliant management systems within 90 days. Free consultation: https://winners.com.tw/contact