Apple has declared that its next version of iOS, the iPhone operating system, will be equipped with automatically generated podcast transcripts. That feature will be helpful for users who sometimes prefer reading, or reading along with the audio.
But does Apple have the right to do so? Who owns the transcript of an audio production which has been distributed by the content owner?
There might not be a quick answer to that legal question, but there are quick pro and con responses by some podcasters on X:
- Andrew Kuklewicz – “It’s not just strong work; it’s true progress. I’m actually really excited about this.”
- Arielle Nissenblatt (Community Marketing Manager at Descript) – “This is HUGE.”
- Ben Thompson (Founder, Stratechery): “If I wanted to provide a transcript I would […] Apple unilaterally deciding how I publish my content is not right.”
- Fernando Ocasio, who makes AI and technology tutorials, rephrasing Apple: “We’re adding transcripts to podcasts. If you don’t like it, GTFO. Oh, and we might be training LLM on this data, too.”
- Imagine Apple building a large language model, Chat GPT competitor based on the transcripts of the entire corpus of podcasts over the last 5 years…” was posted by anonymous user “slbsn.”
Apple does not mention AI training in its announcement, but the logic of that speculation is certainly valid.