Earlier this year, the Reuters Institute asked 264 digital media leaders how newsrooms are supposed to survive the age of AI. The answers boiled down to two clear camps.
First, publishers should double down on work that requires a human body in the room: deep investigations, ground reporting, sharp opinions, and human perspective. They should abandon the stuff a chatbot can churn out in seconds: evergreen explainers, product reviews, generic travel guides, and basic news aggregation.
The second half of the advice is hard to give, but even harder to receive: let it go. Much of that automated content was built on reporting that humans did first, and the AI learned it without paying a dime.
That grievance is the epicenter of today’s media wars. Publishers are suing AI developers, striking licensing deals, blocking web crawlers, and watching search traffic flatline as algorithms deliver answers without a click. The stakes are real, and the anger is justified. But in shape and form, it is an argument Rome already had two thousand years ago, and Roman writers lost it. What they did next is actually a masterclass for publishers today.
A feed anyone could copy
Starting in 59 BC, Rome posted a daily public record in the Forum: the acta diurna. It carried senate business, edicts, trial dates, births, deaths, bizarre prodigies, and gladiatorial results. Scribes copied it, enterprising freedmen compiled the bundles, and businessmen shipped them to senators serving abroad. By Nero’s reign, this gazette had a reach any modern media mogul would envy. The historian Tacitus notes that people complained the acta were read avidly across the entire empire and its far-flung armies, offering it as proof of just how obsessed the Roman public was.
Nobody owned this feed. The writers got a basic scribe’s wage, while the actual profits went to the copyists and couriers. It was general news reproduction, and it was already free.
The most revealing proof of this is a pair of letters from 51 BC. Cicero had been dispatched to govern Cilicia and asked his friend Caelius Rufus to keep him in the loop on Roman gossip. Caelius wrote back proudly saying he had hired a guy to compile everything happening in the city, every decree, rumor, and scandal, and begged Cicero not to be annoyed by the sheer volume. He had literally outsourced the news feed.
Cicero’s reply reads like a modern subscriber furious about a bad newsletter. Did Caelius really think he wanted a list of gladiator pairings, adjourned court dates, and local burglaries? The sort of trivia nobody in Rome would dare bore him with in person?
Cicero didn’t want a record of what had happened. He wanted Caelius to write about what was going to happen, and what Caelius personally made of it. The compiled acta weren’t worth the parchment they were written on. A smart letter from a guy hanging out in the Senate was worth sending a courier across the empire for.
In other words, the Reuters Institute’s modern finding was discovered by an ancient consumer two millennia early. The feed was cheap because it was reproducible. Judgment was expensive because it wasn’t.
What the historian left out
Ancient producers drew this exact line, too. In the thirteenth book of his Annals, Tacitus casually stops the narrative to explain why he refuses to waste ink describing the foundations of an amphitheater Nero was building. Stuff like that, he sniffs, belongs in the daily gazette. Real history is reserved for matters of consequence.
He was actively defining his product against the commodity. The acta recorded everything that occurred. Tacitus selected, diagnosed causes, judged character, and told his readers why an event mattered. He could afford to ignore the trivial because completeness was already available for free in the Forum. The value he sold was omission, and the supreme authority to decide what actually mattered.
If you read the Reuters list of what newsrooms should double down on, it describes Tacitus. If you read the list of what they should abandon, it describes the acta. The survey respondents are essentially begging publishers to become the historian and stop playing the gazette. Tacitus would have gently broken the news that the gazette was never going to pay the bills anyway.
The kidnapped poems
The copyright question itself has an ancient witness: Martial. His sharp, witty epigrams were distributed by mass copying because every Roman book was. Booksellers in the Argiletum sold his work for a few coins a volume and paid him zero royalties. Martial routinely complained that his pages were being thumbed all the way out in Britain but his wallet had no notice of it. There was no legal remedy, and he knew it.
What he could do was police attribution. When a guy named Fidentinus started reciting Martial’s poems as his own, Martial gave the world a brand-new word for it. He called Fidentinus a plagiarius, literally a kidnapper of free-born children, and the word has stuck ever since. His advice to the thief was brutally practical: If you want the poems to be called yours, buy them, and then they will stop being mine. Elsewhere, he noted that a book already famous in the public square can never change its true master.
Martial conceded that the text itself was free. But he kept his name. His readers in Britain weren’t sending him royalties, but they knew whose jokes they were reading, and that reputation was the one thing copyists couldn’t fake or steal. It funded the patronage system he lived on. A voice that a reader recognizes and seeks out will survive a distribution system that pays creators nothing.
The concession that worked
Put these pieces together, and Rome offers a sequence rather than a moral lesson:
The distribution system commoditized the record (the acta were free).
The rights argument was a dead end, and everyone who mattered knew it.
The writers who survived shifted their value to the two things copyists couldn’t reproduce: judgment about what the record meant, and a trusted name attached to the insight.
Publishers today are fighting tooth and nail for the acta. Lawsuits and licensing deals might claw back some cash, and they are worth pursuing, but they are a battle over the reproducible layer, and the reproducible layer was cheap long before algorithms arrived. Cicero wasn’t paying Caelius for the gazette; he was paying for a brilliant correspondent.
The question for a modern newsroom is the exact one Cicero put to Caelius: If your reader already has the feed, what would they still pay a courier for?