AI Learned From Us - Should Creators Get Paid?
There's an uncomfortable truth underneath the AI boom.
Large language models were built on an enormous corpus of human-created material — writing, journalism, books, academic research, code, and more. The intellectual output of millions of people, is now the bedrock for the most valuable technology ever built. In most cases, the people whose work contributed to that value had no say in the matter and received zero compensation or credit.
The ethics deserve a plain conversation:
If creators' work materially improved commercial AI products worth hundreds of billions of dollars, should creators share in that upside?
This question touches virtually everyone — writers, researchers, academics, developers, artists, and any business whose work has been digitized and published online, even if that work was copyrighted.
The legal battles are starting to brew:
In December 2023, The New York Times sued Microsoft and OpenAI, alleging that millions of its articles were used without permission to train large language models that now compete directly with the Times as a source of information. The case is ongoing. Eight additional newspapers — including the Chicago Tribune and the New York Daily News — filed a similar suit.
Sony Music, Universal Music Group, and Warner Records sued AI song-generation startups Suno and Udio in 2024, alleging copyright infringement in their training data.
In September 2025, Anthropic agreed to a proposed $1.5 billion settlement to a group of authors Andrea Bartz, Charles Graeber, and Krik Wallace Johnson after they filed a class-action lawsuit against the company for using millions of copyrighted books - including ones on pirate sites - to train its AI models. The settlement is widely viewed as a significant signal that the use of unlicensed copyrighted works in AI training carries substantial legal and financial risk.
Even in jurisdictions where "fair use" arguments hold up legally, the ethical question remains open. AI companies defend their training practices by arguing that using data to learn from is fundamentally different from copying and distributing content. That argument has found some legal traction. But it sidesteps a harder one: if the patterns, knowledge, and voice encoded in a model came from human creators, who gets credit for the original patterns?
From a philisophical standpoint, the answer seems clear. The creators should be credited. More than credited — they should have had the opportunity to consent, to opt in or out of having their work ingested. Now that these models have already been trained on vast swaths of the internet, opting out retroactively is nearly impossible. Which is precisely why compensation frameworks matter: not just as a retroactive moral acknowledgment, but as a precedent that protects future creative work.
The approaches most commonly discussed in policy conversations include:
Collective licensing / rights pools — similar to how ASCAP and BMI have long managed music royalties
Opt-in / opt-out regimes for training data, giving creators meaningful control going forward
Attribution standards when AI outputs can be traced to high-confidence sources
Revenue-sharing or data dividend models — direct compensation tied to how much a creator's work contributed to model training
For companies using AI throughout their business functions, this isn't just a philosophical debate. It's a reputational and legal risk question that belongs in your AI governance conversations right now.
The businesses most exposed are those that:
Use AI-generated content without disclosure
Rely on AI tools whose training data sourcing is opaque or legally questionable
Haven't audited the AI platforms they use for compliance with emerging standards
The New York Times lawsuit alone illustrates how significant the financial stakes can be — and courts are beginning to issue rulings that will define what's permissible and what isn't.
What This Means for Business
Approach AI-generated content the same way you'd approach using someone else's research or creative work. Cite your sources where possible. Vet the AI tools you use for transparency about their training data. Prefer platforms that have established licensing agreements with content creators. And watch the legislative landscape closely.
AI can absolutely be a powerful partner to creators and businesses alike. But partnership implies consent and fair value exchange. The creative work that made these models possible deserves recognition. And as the legal landscape continues to evolve, the companies that treat this as a values question — not just a compliance question — will be better positioned when the rules fully arrive.
Do you have thoughts on AI ethics? I'd love to hear them! Email me at Elizabeth@Elizabeth-Marketing.com
