Musicians are being “squeezed out” and “marginalised” from their industry by AI, researchers have said. A study from Queen’s University Belfast warns that consumers and musicians are being “shortchanged” because of lack of controls over the technology.
Research Highlights Malicious Use of AI
Dr Katherine Pittalis has examined how AI is being used in music-streaming platforms, online retail for concert tickets and music-related content on social media. She argues that the UK music regulatory framework is “woefully lagging behind” that of the EU and US in terms transparency around malicious actors’ use of bot technologies.
Her research examines phenomena including ‘metric bots’, which can inflate metrics such as likes and follows on social media, along with ‘artificial musicians’, the combination of deepfake technology with bots to create an AI ‘influencer’. The malicious use of technology is becoming increasingly sophisticated, according to the research, with bots ever evolving to circumvent the rules laid down by the social media, music streaming and primary ticketing platforms.
Scams and Unregulated Platforms
Scams are one example highlighted of the increasing exploitation of technology, particularly on ‘secondary market platforms’, where tickets for high-demand concerts are sold at inflated prices. The research identified multiple instances where consumers handed over large sums of money only to find the tickets never existed.
The research also found that more informal ‘buy and sell’ groups, such as those that individuals can set up on Facebook, are “highly unregulated” and even, on occasion, moderated by bots. Elsewhere, it found instances of bots moderating ‘community notes’ sections on social media – despite the fact that community notes were established explicitly to combat misinformation and fake news.
Call for Urgent Regulatory Action
Ms Pittalis said the study paints “quite a depressing picture”. “Social media platforms are incredibly secretive about the use of bots,” she said. “One study I consulted even found that X (formerly Twitter) has more bots now than when Elon Musk took it over and promised to get rid of them. From my research, the bots I discovered look more authentic in how they behave and interact with organic users. Even with protocols and safety measures in place, bots are continuing to thrive in these online environments and distort organic users’ perception of what is real and what is misleading content that has the potential to cause harm.”
Ms Pittalis is calling for the UK’s music industry, social media companies and policy makers to tighten up cyber-security urgently to ensure a better deal for consumers and artists. “It’s not fair,” she said. “Consumers and authentic musicians are being squeezed out, marginalised by the very industry meant to support them. It might sound apocalyptic, but the truth is the bots are running amok – UK legislation has fallen behind. There’s no anticipation of future trends. Instead, it’s reacting to what’s already happened. It needs to be future-proofed, which includes more oversight into how social media and other online platforms control breaches and uphold user-safety from bots. The industry needs to take this seriously before it’s too late.”
Further Research and Industry Response
For a book stemming from the research Ms Pittalis is also investigating the use of social bots to spread malicious information designed to damage a musician’s reputation, such as the spread of explicit deepfake images of the pop star Taylor Swift on social media platforms in 2024. She will also look at AI-generated music, as musicians start to use aspects of AI in their production, but in some cases entire songs are being generated entirely by AI. There is a growing demand for this type of music to carry a transparency label but UK policy is still in development. Spotify last week announced that from mid-September, ‘AI Persona’ badges will start to appear on artist profiles generated by AI.
The research was supervised by Professor Giancarlo Frosio from the School of Law and Professor Hans Vandierendonck from the School of EEECS at Queen’s. Mr Frosio said the research “makes an original and timely contribution to our understanding of how malicious automation is reshaping the music ecosystem”. “Its central insight is that this is neither simply a platform problem nor a state problem; effective regulation requires clearer legal duties and enforcement alongside much stronger platform-level detection, transparency and accountability.”
