This illustration picture shows Pope Leo XIV Encyclical Letter “Magnifica Humanitas”, focused on the rise of artificial intelligence, in The Vatican on May 25, 2026. (Photo by Alberto PIZZOLI / AFP via Getty Images)
AFP via Getty Images
Content publishers are fighting a multi-front battle – it might not be hyperbole to call it existential – before courts and regulators to protect their business from the onslaught of AI. Two recent reports – including one from an unexpected “higher authority” in the Vatican – may help shape that ongoing struggle.
The double whammy from large language models (LLMs) like Google Overviews, OpenAI’s ChatGPT, Anthropic’s Claude and Perplexity is the years of unauthorized use of copyrighted materials to help train their models and the subsequent destruction of publisher audiences from publisher material providing the foundation for answers to AI prompts.
For publishers who have long relied on Google Search as a tool to drive awareness and reader engagement, LLMs are wreaking havoc. As I noted back in February, the click-through rate for links embedded in Google Overviews drops 95-96% compared to traditional Google searches. For now, the devasting impacts for media content from LLMs are primarily borne by text-focused publications. But the threats to video will ramp up as LLMs will get better and better at scraping video as well. First, they came for the publishers, and who was there to speak for them?
The Pope Weighs In
I did a check of my Forbes archives and I’m fairly sure that there have been no pontiff references in the last twelve years. But the first encyclical from Pope Leo XIV is likely to be a remarkable contributor to the still-early global debate over the ever-enveloping role of AI in society. The Pope’s Magnifica Humanitas (Magnificent Humanity) isn’t providing a roadmap for stopping AI through law, technology or the will of God. And the encyclical isn’t a purely anti-Big Tech screed. Its likely influence is far more subtle.
The lack of a legal framework doesn’t diminish the thoughtfulness, and one hopes the thought-provoking nature of the Pope’s entreaty to grapple with the transformative nature of AI, including the repercussions for those who create for a living. The Pope writes that for “practical ways of fostering fairness, participation and care for creation must be found.”
In an age of AI slop and ever-growing skepticism about what is and isn’t real, the Pope states that “fidelity to the truth requires integrating the possibilities offered by technology within a framework marked by wisdom, which is capable of safeguarding both the dignity of each person and the future of our common home.” Grappling with these massive challenges is a “shared responsibility” for us all. This isn’t your usual ad tech jargon, but hopefully a guidepost for greater societal engagement on this existential challenge.
It’s Not Just the LLMs
I don’t think Matthew Goldstein quite tops the Pope’s 65 million social media followers, but he did recently produce a report on unauthorized AI content extraction that provides a great deal of meat on the AI bones referenced by the Pope’s encyclical. Goldstein is an experienced advertising industry analyst who has shifted his focus towards AI in recent years.
Goldstein’s report on the “scraper economy” moves past the analysis of the major LLMs to a host of mostly lesser-known companies that do a great deal of the dirty work of scanning the internet for content without licensing it. It’s all about seeking forgiveness – and maybe not even that – rather than permission.
Although the scrapers may be paying nothing to take content, they are certainly getting paid for their efforts. The twenty-one companies that Goldstein’s report profiles collectively have a business that in a noticeably brief time tops $1 billion and grows at double-digits. That money gets collected from a host of legitimate enterprises that are paying these scrapers – but not the creators of the content – to “extract publisher content at industrial scale without compensation” for the content producers. This all sounds a bit like a digital fencing operation – you get me the goods; I’ll get you the cash. For anyone watching Apple TV’s Friends and Neighbors it might feel a little familiar. You only have a business in stealing if people are willing to pay you for what you’ve stolen.
Goldstein describes a multi-layered web of the scraper economy, including foundational companies that make other scraping by others possible, led by a company called Bright Data that the Goldstein report claims generates $300 million in revenue itself. There are a number of “crawl and extraction” companies such as Firecrawl and Grepsr transform the extracted content into formats useful for machine consumption.
Another group of scrapers Goldstein identifies as “the most direct competitor to publishers” provide content to AI models that respond to queries in direct competition to the platforms publishers provide directly to their users. One of these companies, You.com, has a current valuation of $1.5 billion. Not exactly OpenAI dollars, but nothing to sneeze at.
Who Gets to Reap Benefits Here?
Goldstein reveals a host of establishment institutions that regularly deploy scrapers, including OpenAI, Apple, BCG, IBM and the law firm Latham & Watkins (I hate to make the analogy, but even the Mob has lawyers). What is most disturbing, however, is the number of brand name publishers that are using scrapers (of other people’s content) for their own research purposes, a group that includes The New York Times, NBC, BuzzFeed, the BBC and Forbes for their own research purposes. And this is from publishers who in many cases are actually suing the LLMs for the unauthorized use of their own content. What is it again – do as I say, not as I do?
Unlike the Pope, Goldstein is more proscriptive in terms of how publishers should approach this world. There is no such thing as “stopping” AI or LLMs. His solution framework would include making deals with scrapers and getting paid. At the end of the day, it is likely that the only real solution for publishers is a system that pays them fairly based on the usage of their content, whether that emerges from lawsuits, regulation or plain old dealmaking. Is that fair enough? As s tortured a road as that may be, it is still a more promising scenario than hoping technology is going to put a stop to the scraping economy. So much more to this story to follow.