AI is making (all the) headlines today, but a critical aspect often overlooked is how tech companies need existing content to train their AI models—sometimes using that content without the owners’ consent.
AI models, especially those involving machine learning and neural networks, require vast amounts of data to learn and function effectively. This training data, which can include text, images, audio, and video, allows the AI to recognize patterns, make predictions, and generate content.
For example, training a music recognition AI would necessitate thousands of hours of music, including songs, lyrics, and metadata. These models process and learn from this data to improve their performance over time.
Sony’s take
Recently, Sony Music Group addressed the issue head-on and issued a stern warning to AI companies and music streaming platforms about the unauthorized use of its content for AI training.Sony Music Group, one of the leading record labels globally, emphasized that any use of its content – including album cover art, metadata, musical compositions, and lyrics – requires explicit permission. Sony, representing artists such as Lil Nas X and Celine Dion, sent warning letters to over 700 companies to protect its intellectual property from being exploited without consent.
Although the list of companies contacted by Sony was not disclosed, it’s plausible that Apple, a major player in both AI and music streaming, could be among them. Apple, with its extensive Apple Music library, has both the resources and the incentive to use this content for training its AI models. However, unless Apple and Sony have a specific agreement permitting this use, Sony’s warning suggests that such activities would violate their intellectual property rights.
#Sony #waves #red #flag #model #trained #label #content #explicit #permission
