The Napster Moment For Journalism & Publishing

$1.6 billion.

That's what ASCAP alone distributed to songwriters last year. Not the entire music industry. Just one organisation. In one country.

Publishing's entire collective licensing infrastructure was $212 million in the same year.
Barely 11% of ASCAP’s single-year total.

Music collects more in 48 hours than publishing manages all year.

Here's the kick in the proverbials:
Anthropic would rather pay billions in settlements than deal with your licensing system. They literally chose to risk catastrophic legal liability than navigate your infrastructure.

“Business slog”, they called it.

Sit with that for a second.

This happened because music learned one rule publishing never did:
one deal beats a thousand contracts.

While you were sending DMCA takedowns, OpenAI scraped 3 million articles from The Washington Post. While you debated digital rights, Meta trained on Books3 – 170,000 pirated books they grabbed from a torrenting site.

Anthropic downloaded 7 million titles from LibGen and Z-Library. Russian pirate sites. They knew they were stealing, then budgeted for the lawsuit, and did it anyway.

But why?
Because stealing from publishers is literally more cost-effective than licensing.

Judge Alsup proved this in Authors Guild v. Anthropic case. He ruled that training on legally-obtained books is "fair use", BUT that Anthropic's pirated millions weren't legally obtained. Faced with a December trial, they instead settled for $1.5 billion in August.

Think about that:
A billion-dollar AI company considered crime more convenient than compliance.

They found it easier to pay billions in damages than to navigate thousands of fragmented publishing licenses.

Some might say it’s moral failure. I suggest that’s market failure.

Remember Napster? Music faced freeloading teenagers with DSL connections sharing MP3s with each other.

You're facing companies worth more than Sweden's GDP with data centres the size of small cities.

But here's what nobody understands: music didn't win with lawsuits. They won with better infrastructure and incentives.

The solution was iTunes first, later Spotify. And everyone bought into the idea.

ASCAP and BMI turned millions of songs into one license. This made compliance easier than theft. Today, 95% of music creators and distributors participate in collective licensing.

In publishing participation sits at 6%.

Ninety-four percent of publishers aren't even in the game.

The UK's Publishers Licensing Services (PLS) is actually MORE efficient than ASCAP, distributing 94% of revenues versus ASCAP's 90%. But talking about efficiency at 6% participation is useless.

Efficiency is not the problem. Scale is. And buy-in.

The most efficient collection system in publishing covers fewer than one in twenty publishers. You can’t manage a digital economy with paperclip infrastructure.

Music had chokepoints: radio stations, streaming platforms, live venues. You could count them, license them, sue them if needed.

Publishing has no chokepoints. Only fragments. Articles become paragraphs. Paragraphs become sentences. Sentences become AI training data.

AI doesn’t “copy”, it consumes. Billions of snippets, stripped of metadata, blended into models.
You can’t chase that with traditional licensing contracts.

The bit that should make you furious is every morning these companies wake up and choose crime. Not because they're evil. Because the system makes crime convenient.

Music fought Napster with alternatives that were easier to use, and felt almost free. Publishers have circumvention pretty much built into their code.

It’s not because they’re evil, it’s because the alternative is much easier.
They want to avoid “business slog”.

Here’s the part nobody’s saying out loud:
This isn’t only about defence, it’s about intelligence.

Attribution and royalties.

Music PROs don’t simply collect money; they collect data. And they use it.

They know who’s trending, where, when, and with whom. That insight fuels the next hit before the market even notices.

Imagine knowing:

The future of publishing, then, changes from protection to prediction.

Blockchain.

Yeah, I said it.

But forget crypto hype or NFT nonsense. This is about making ownership undeniable and compliance automatic.

When ownership embeds in content itself, not some database that pirates ignore, the incentives change. Smart contracts execute licensing automatically. Payment happens instantly. Suddenly, stealing becomes much harder than simply paying.

The global text-mining market is £670 million and growing 23% annually. That's just what's tracked. The real number including dark usage? Multiply by ten.

Publishers who build this infrastructure first become the market, rather than simply participants. Everyone else gets commoditised into whatever terms you set.

The window is maybe 18 months before AI companies lock in their preferred sources.

After that, your content either flows through your pipes… or someone else’s.

Seems like there’s three options:

This isn't about IF you're being robbed. It's about how much they're taking.

The Anthropic case proved it: AI companies would rather pay billions than deal with licensing chaos.

So build the infrastructure that makes licensing simpler than stealing.

The tech is ready.
The timing is perfect.
The window is closing.

One deal beats a thousand contracts.

The question is whether you'll be making that deal or living under terms someone else negotiated.

-John

This is where you can get advice, insights, and learn tips and tactics to build a better business in journalism or publishing, whether you’re a freelance, an indie publisher or working in a larger organisation.

We’ll cover everything from business strategy and commercial focus, to technology, legal and more.

If you're already looking at licensing infrastructure, book a quick call. 15 minutes could save you months of dead-end strategy.

If you're still exploring, I’ve distilled Google’s 48-page ROI of AI guide into a 1-page cheatsheet with the most relevant insights specific to publishers.

Grab it below. (Find Google’s full report here.)

See you in the next edition!