The issue of copyright and artificial intelligence (AI) is highly topical. Three years after the launch and rapid expansion of artifi- cial intelligence based on large language models for mass use (November 2022) – such as ChatGPT – the question of who owns the copyright to AI-generated content remains unresolved. Another major issue, closely related to copyright in AI-generated content, con- cerns the use of the voices of real individuals and creators – namely, the unauthor- ized use of voices by artificial intelligence. Is legal protection against AI-generated deepfakes and voice cloning possible, and can creators and public figures control how their digital identity and voice are used by AI? The training of AI models also raises important questions: does the use of cop- yrighted materials for AI training constitute copyright infringement, or are AI sys- tems trained solely on public domain materials? What legal consequences should follow in these different scenarios? This article examines and discusses some of the initial problems and cases arising from the use of AI in the context of copyright law.