The UK’s Competition and Markets Authority (CMA) has launched an initial review of “AI foundational models,” such as large language models or LLMs powering OpenAI’s ChatGPT and Microsoft’s Bing Chat.
The CMA’s investigation will focus on the competition and consumer protection considerations of developing and using these AI models, and could also cover generative AI models that power AI art platforms like OpenAI’s DALL-E or Midjourney.
The CMA’s review will focus on competition and consumer protection considerations in the development and use of AI foundational models, the authority said in a notice on Thursday (May 4).
The review follows the release of a white paper by the UK government in March, asking regulators, including the CMA, to consider how the innovative development and deployment of AI can be supported across five overarching principles: safety, security and robustness; appropriate transparency and explainability; fairness; accountability and governance; and contestability and redress.
“The review seeks to understand how foundation models are developing and produce an assessment of the conditions and principles that will best guide the development of foundation models and their use in the future.”
UK’s Competition and Markets Authority
“In line with government’s AI white paper and the CMA’s role to support open, competitive markets, the review seeks to understand how foundation models are developing and produce an assessment of the conditions and principles that will best guide the development of foundation models and their use in the future,” the CMA said.
The term “foundation models” were coined by Stanford University’s Human-Centered Artificial Intelligence Center’s Center for Research on Foundation Models in 2021. These models are AI systems that focus on training one model on a wide range of data and adapting it to many applications.
Foundation models are poised to transform much of what people and businesses do, including in the music industry. They have the potential to generate lyrics, compose melodies and create entire songs.
Last week, Los Angeles-based venture capital firm Sound Ventures closed its new Sound Ventures AI Fund at nearly $240 million, with plans to invest in foundation models and other similar natural language processing technologies capable of generating human-like responses to natural language inputs.
“We believe this is potentially the most significant technology we will experience since the advent of the internet,” actor Ashton Kutcher, who also is a General Partner at Sound Ventures, said.
Apart from the ChatGPT chatbot, OpenAI also has a music-related generative model called Jukebox that can compose original music in a range of styles including pop, country, reggae, rock, jazz and R&B.
In 2019, TikTok owner ByteDance bought UK-based AI music startup called Jukedeck, which creates royalty-free music for user-generated online videos, while stock photography company Shutterstock acquired AI-driven music platform Amper Music in late 2020.
However, the use of foundation music in music also raises concerns about copyright and ownership.
The recent “fake Drake” and AI-generated music controversy recently sparked a debate in the music industry after a mysterious ‘artist’ called ghostwriter released a track featuring AI-replicated Drake ‘vocals’ that racked up hundreds of thousands of streams online before they were pulled down.
The copyright issue is among the concerns that prompted the CMA to open a probe into generative AI.
The CMA is seeking views and evidence from stakeholders until June 2. It will publish its findings in September.
Music Business Worldwide