Automated-answer systems that ingest, store, or reproduce copyrighted material can unintentionally infringe copyright if they output verbatim excerpts, replicate distinctive creative elements, or produce derivative works that are substantially similar to the original. Even when a system merely references an artist’s work (summaries, metadata, or short quotations), the safe legal footing depends on jurisdiction-specific doctrines such as fair use/fair dealing, the amount and character of the excerpted material, and the purpose of the use (transformative, commercial, educational, etc.).