Opted In by Staying
On the thirtieth of October, every newsroom with a Facebook page will agree to let Meta read its website, and there's no button to refuse.
0:00 / 12:14
Opted In by Staying
Meta starts reading news sites by default on 30 October. In the same week, Google said more than 400 publications are now being paid for their journalism. That is the week's question in two sentences: who asks before reading you, and what does the answer cost. What follows is the detail behind each answer, in the order the briefing takes them.
The default: Meta's new commercial terms. Category: AI and platform. Assessment: trend to watch, with an action that falls due before the deadline.
From 30 October, the business terms for Meta's products authorise automated software — scrapers, bots, crawlers — to access, index, cache and collect any domain or web page "associated with" an account holder's products or services, and to keep pulling updates from it. There is no option for a business account to opt out before the effective date. 1
After Nieman Lab reported on 15 October that the draft also permitted training use, Meta revised the terms the next afternoon. The word "train" was removed and the permitted uses narrowed to "advertising and business products." 2
Three lawyers read the revision and none of them was reassured. James Grimmelmann of Cornell Law School said the remaining language is "still broad enough to drive a truck through," because "develop" can include "train" and "advertising and business products" is not a defined term. Matt Topic of Loevy & Loevy said publishers should remain concerned. Benjamin Sobel of the University of Wisconsin compared robots.txt to a homeowner's sign reading "Don't look at my house": the preference can be expressed, and it can still be ignored. 1
Meta's position is that nothing has changed: a spokesperson said the revision was meant to make clear that robots.txt preferences are respected, and that sites can switch the crawler off with those flags as they always could. No such promise appears in the terms themselves.
The trade response was immediate. The UK News Media Association called it a land grab; the News/Media Alliance in the United States called it "outrageous and probably illegal"; the European Publishers Council said it would ask the European Commission to scrutinise the terms. On the same day, eight UK national newspaper editors and chief executives signed a letter urging the Prime Minister to back the Automated Online Software (Access and Transparency) Bill, which would require bots to declare who operates them and what they intend to do with the content. 3
What a single newsroom can do about it does not need a lawyer:
- Establish which parts of the organisation hold business accounts. The terms bind the account holder, and in most organisations that is marketing, not the editor.
- Read the robots file at the site root and check what it actually says about the named AI crawlers, since that is the control Meta says it honours.
- Record the decision and its date, so the same argument does not restart from nothing next time.
The paid route: Google's News AI pilot. Category: Revenue and AI. Assessment: an immediate opportunity for publishers who can get in, with the caveat that the money is undisclosed.
Google said on 15 October that more than 400 publications worldwide now take part in the pilot it began last year with partners including The Washington Post and The Washington Examiner. The new cohort runs from CNN, the Financial Times, Axios and Forbes to local publishers: The Post and Courier in Charleston, the Texas Tribune, Oil City News in Wyoming and Shaw Media, which owns titles across Illinois. 4
Google pays for what it calls the delivery and use of enhanced content. The features are article overviews on Google News publisher pages, and audio briefings a listener can customise. Attribution and direct links back to the source are part of the arrangement. Financial terms were not disclosed. 5
The enforcement route: Reach. Category: Revenue and AI. Assessment: trend to watch.
Reach, which owns the Daily Express, the Daily Star and OK!, has signed a Google deal that Digiday understands runs two years. Its chief executive, Piers North, described the value as setting the rules rather than replacing lost referrals: content cannot be used without permission, and the deal is time-limited, so both sides learn something to measure at renewal. 6
Reach also sent OpenAI a cease-and-desist letter. North said the group's access logs show well over 100 million successful scrapes of its content in six months, and that the scrapers have become hard to distinguish from readers.
The collective route: SPUR. Category: Revenue and AI. Assessment: an immediate opportunity for small publishers, at the price of a membership.
The coalition's full name is Standards for Publisher Usage Rights. It launched in February with the Guardian, the Financial Times, the Telegraph, the BBC and Sky News. This month it added more than 40 members — among them The New York Times, The Atlantic, Agence France-Presse, the hyperlocal Blog Preston and the Independent Community News Network, which represents 120 independent and community titles — for a total of 80 organisations from more than 20 countries. 7
Earlier in October the coalition published version 1.0 of an open-source content telemetry standard. It gives AI companies a common format for recording what they scraped, what they passed to a model, how they cited it and whether a reader clicked, and for reporting that back to the publisher. 8
The reader route: London Centric. Category: Subscriber conversion. Assessment: an immediate opportunity, and the cheapest one here.
Jim Waterson launched the London newsletter in September 2024 after leaving the Guardian, where he was media editor. It now has 5,500 paying subscribers and more than 42,000 free ones, and it is the biggest paid local newsletter on Substack in the UK. 9
Three people work on it full time: Waterson, a reporter and a head of growth hired in July. Revenue is between £300,000 and £400,000 a year, and about 95 percent of it comes from subscriptions. It has never spent money on marketing. Substack takes a 10 percent commission on subscriptions; the publication also earns a little from syndication.
Waterson puts the growth down to two things. His WhatsApp number sits in every piece and produces 20 to 30 tips a day that could be national stories, and almost all of the best stories now come from readers. The second is the bar: every edition has to be an eight or nine out of ten, because readers will pay for a scoop about the city they live in and not for coverage where the effort is missing.
The negative result is as useful as the positive one. Instagram is near 95,000 followers and YouTube has more than 350,000 views, and Waterson calls the earnings abysmal. Prices have not moved since launch: £60 for the first year, then £79.50. He describes the newsletter as the local add-on to a Times or Financial Times subscription.
The method: a lab that tests the question for you. Category: Audience growth and research. Assessment: an immediate opportunity, at the cost of two or three conversations.
Local NExT Lab at Syracuse University, directed by Joshua Darr, published its first year's report this month. Mississippi Today asked whether the stories published around its Pulitzer-winning investigation changed how the investigation was received. The lab showed the same investigation to 1,768 adults, framed three ways. 10
Readers rated the investigation more accurate, more important and more relevant when it appeared beside community stories. A mix of legislation coverage produced no consistent result. A second experiment found that Instagram Reels featuring a reporter's face drew the most interest, but that when people chose to follow, accounts that looked like a news brand beat accounts named after a reporter.
The lab's argument is that dashboards show movement and not cause: "Engagement data can't tell you what caused audiences to engage, settle a hunch, or isolate your contribution." Randomisation can. Newsrooms contribute two or three conversations; the researchers design, field and analyse the test.
The decision this week is small and it has a date on it. Before 30 October, find out which accounts your organisation has consented with, read what your robots file actually tells the crawlers, and write down what you decided. The licence, the letter and the coalition are all arguments about the price of the same journalism. The readers paying £60 a year are the evidence that it can still be sold.
References
- 1
- 2
- 3Guardian and Sun agree on ban for deceptive AI bots
pressgazette.co.uk
- 4
- 5
- 6
- 7New York Times among surge of sign-ups to publisher AI coalition SPUR
pressgazette.co.uk
- 8
- 9
- 10
This story was produced automatically by a channel. One sentence is all it takes for Neodrop to keep producing for you.