Artificial intelligence (AI) is rapidly making its way into creative industries, sparking both fascination and fear among artists and musicians. AI’s potential is immense, yet it also poses significant challenges related to intellectual property, consent, and the essence of human creativity. As AI-generated songs and illustrations increase, we face a new reality that scholars like Oliver Bown from the University of New South Wales have termed, “The robot soundscape.”

The integration of AI in music production once seemed exciting. Musicians and composers were curious about AI’s ability to create new sounds and compositions, potentially reducing barriers for emerging artists. However, there are growing concerns about copyright infringement and the ethical use of AI-generated content, mainly when it involves works by living artists.

For example, TikTok user Ghostwriter977 released a song titled “Heart on My Sleeve” on April 4, 2023. The song employed AI voice filters to replicate the voices of Drake and The Weeknd with such accuracy that some listeners believed it was a new collaboration between the two artists. Consequently, this caused significant discussion about the ethical implications of the project. Later, Drake and The Weeknd both stated they did not consent to this.

Last year, in a historic move, the Recording Academy announced that only human creators are eligible for the Grammy Awards, explicitly stating that, “a work without human authorship is not eligible in any category.” Separately, earlier this year, Universal Music Group (UMG) and electronic instrument maker Roland Corporation teamed up to create a set of guidelines that the companies published under Principles for Music Creation with AI.

The manifesto establishes seven basic principles:

• Music is fundamental for humanity;
• Humanity and music are inseparable; Technology has long supported human artistic expression and, applied sustainably, AI will amplify human creativity;
• Works created by humans should be respected and protected;
• Transparency is essential for responsible and trustworthy AI;
• The perspectives of musical artists, composers, and other creators should be sought and respected;
• And pride in helping bring music to life.

Both the Grammy’s decision and that of UMG and Roland Corporation underscore the tension between the potential of AI and the need to preserve human artistic integrity.

However, the influence of AI extends beyond music. Generative AI tools like ChatGPT and Stable Diffusion train through extensive datasets that may contain copyrighted material. While these tools do not directly replicate, they can learn and produce results that violate copyright protections.

“Generative AI uses deep learning to understand the relationships between data points, such as which brushstrokes tend to follow others in a painting style,” said Artificial Intelligence and Machine Learning coordinator at Fanshawe College, Jim Cooper. “This allows AI to create images that can be remarkably similar to copyrighted works.”

The core of the debate around AI in the arts revolves around intellectual property and consent. Sony Music Entertainment, Universal Music Group Recordings, and Warner Records Inc. have all filed copyright infringement lawsuits against AI music generation services Suno and Udio. These lawsuits raise questions about whether using copyrighted materials to train AI constitutes infringement and whether there should be a “fair use” exemption to copyright law.

“Copyright infringement is a major ethical issue with large language models and generative AI,” Cooper noted. “These models consume billions of words, images, and sounds, many of which are protected by copyright. The challenge is ensuring that AI-generated content doesn’t violate these protections.”

Addressing these challenges requires a multi-faceted approach. Policymakers and regulators must establish clear guidelines to ensure that AI-generated content respects intellectual property rights. Companies developing these tools must also implement robust filtering mechanisms to prevent copyright infringement.

“There are no established methods for building guardrails into generative AI yet,” Cooper said. “However, companies could restrict model outputs and use filtering to limit copyright infringement. Regulation may also be necessary to ensure compliance and protect creators’ rights.”

It looks like the music industry is already taking steps in this direction. YouTube has begun talks with Sony, Universal and Warner to licence music for AI tools.

For AI to positively impact the arts, consent and collaboration must be the primary focus. Artists such as Grimes and Killy have openly allowed their voices to be used in AI-generated music, highlighting the significance of consent in this new creative environment.

The rise of AI-generated music and art also demands we rethink traditional notions of authorship and creativity. While AI can assist in the creative process, human input remains essential. As Cooper put it, “AI is a tool that can enhance creativity but cannot replace the human touch.”