Rules needed for AI-related copyright violation

In most modern societies a strong mechanism exists to protect people's intellectual property rights. While copying or publishing content created by someone else, be it text, art or video, will surely invite charges of plagiarism, those who read someone else's creative idea and create a similar story without giving due credit to the original author also risk being sued and made to pay compensation.

However, what happens if the entity that copied or reproduced content was an artificial intelligence? What if an AI model makes use of the large amount of content it is fed from online sources to write a story? Does that also amount to copyright infringement?

The lawsuit filed by The New York Times against OpenAI and the agreement between Apple and National Broadcast Company have put the spotlight on these questions. While the NYT claimed OpenAI's "unlawful copying and use of The Times' uniquely valuable works" led to "billions of dollars in statutory and actual damages", Apple got NBC's approval for collecting their data by paying $50 million, both of which lead to the same conclusion that data is money and the author's right should be respected and paid for if someone wants to profit from it.

That being the basic principle, how to divide interests between the original writer and the AI developer and how should one side pay the other? After all, AI is a tool created by human intelligence to make life convenient for everyone, not just the AI developer. If the profit margin for AI developers dims, the emerging industry might find it difficult to sustain itself. However, if it cuts into the interests of original writers, reporters, composers and other creators, people might be discouraged from creating content in the first place.

In a deeper sense, the NYT versus OpenAI case is very relevant to the times we live in. The lawsuit will definitely not be the last one, but it will help decide how to share interests between the original writer and the AI developer.