Keeping a page on the open web costs money, even when anyone can read it for free. Someone pays to keep the servers running, answer requests and send another copy to a reader who will never receive an invoice.
In July 2026, Lane Becker made the case for Wikimedia Enterprise. The service charges business clients to deliver Wikipedia content. The content keeps its open license. Clients pay for a feed they can rely on, at the speed and scale they need. The fees help cover costs that donors would otherwise pay. Wikimedia also lists free access routes for smaller users and other groups that qualify.
A company that builds a product from a public resource should help pay to keep that resource open. The danger comes when we stop asking what remains available to everyone outside that bargain. That is where the privatization of the open web becomes a serious concern: in the public route that shrinks, gets harder to use or shuts down after the private route has found its customers.
Charging for access is not one thing
A paywall changes what a reader can see. A limit on requests from bots changes how fast a machine can gather pages. A license sets terms for how the work can be used. An exclusive content deal, where one exists, can give one buyer rights that others cannot get. These rules can overlap, but they do different work.
We lose sight of the costs and rights at stake when we treat all these rules as the same thing. A site can remain free to read while blocking a bot that makes millions of costly requests. It can charge for a feed that saves time while letting someone else gather the same openly licensed work through a slower public route. Neither choice, by itself, takes the work out of public life.
Becker’s case deserves a fair hearing. Heavy business use can drive up costs for a group whose mission is to serve everyone. If it has to choose between charging a heavy user and asking ordinary donors to cover that user’s bill, charging looks reasonable. Better delivery can also be worth buying. Keeping a feed working, helping clients and sending prompt updates take work even when the facts and words are free to use.
A buyer might pay for a faster service or technical help. It might pay for permission to use the content, perhaps on terms that exclude other buyers. Those terms deserve scrutiny even when the announcement calls the deal a “partnership.”
Nor does every publisher owe every commercial crawler unlimited service. A small news site may need paying readers to survive. An archive may have servers that cannot cope with heavy use. Keeping a public route open can include reasonable limits. It cannot mean giving away someone else’s labor and bandwidth with no end in sight because we like the sound of an open web.
Anyone who argues for an open web needs to account for those costs. Otherwise, we risk asking the public to pay the bills of whichever company can send the most requests.
Who gets left outside a private deal
A provider that needs money can reach terms with a large buyer that needs dependable access. Both can accurately describe the result as a success.
The person who wanted to build a much smaller service was not part of the talks. Neither was the researcher whose project has no staff to work out a deal. The deal may protect their access. An open license or a separate access program may do the same. They may also be overlooked. A glowing press release tells us very little about that.
A publisher may earn more money from a deal while fewer people can use its work. The publisher benefits, but we still have to ask what happened to everyone else’s access.
Legal scholar James Grimmelmann’s work on AI scraping warns of harm from the barriers sites put up. Rules meant to stop unwanted bot use can also get in the way of reading, linking and saving copies of pages. The warning does not prove that every rule has caused such harm. It is still a reason to examine who gets caught at the gate.
A barrier seldom arrives bearing the name of the worthwhile project it will prevent. It arrives as a sensible answer to an immediate problem: too much traffic, too little revenue, an unwanted use. Each decision can make sense within the institution making it. Taken together, they can leave a web that is easy for an established company to buy its way through and difficult for anyone else to explore.
The worry is sharper when someone must approve each request for access. A public page explaining the conditions gives a small operator something to work with. A form asking them to contact the partnerships team leaves them waiting for someone to respond. Neither route involves a reader paywall, yet only one lets the small operator get started without that reply.
We should be careful about what we infer. Keeping the terms private does not prove that only one buyer has rights. A paid feed does not prove that free access has been lost. A blocked crawler does not prove that human readers have been shut out. The criticism needs evidence about the route that has changed. Without it, “privatization” becomes a mood attached to any exchange of money.
But the reverse mistake is just as easy. A page can remain visible in a browser while people lose the means to study, save or build with a large body of work. There is more to public access than being able to read one page at a time.
Keep a public route open
A deal worth signing should spell out what remains open to everyone else.
For openly licensed work, that means respecting the rights people already have and keeping ways for them to use those rights. For material that has never carried those permissions, it means being clear about what is offered, what is restricted and why. We cannot ask the same of Wikipedia, a small paid news site and a database run for profit. The question can stay the same even when the answers differ.
The access terms should answer practical questions:
- Can a person find the access rules without knowing someone inside the organization?
- Are the formats useful?
- Can a small project work within posted limits, or must it ask for a deal before it can do anything useful?
- If access is withdrawn, is there an explanation and a way to correct a mistake?
An institution can publish those terms while charging for a costly commercial service. People using the public route should be able to tell what they can do and what happens when something goes wrong.
There is also a funding question for those of us who want that route to survive. Public access takes work. The groups that provide it need funds, and fees from business users can help. A promise without a budget may not last through the next surge of requests or the next round of cuts. The funding should help keep access open, with someone charged with doing the work.
The useful test of a deal comes after the major customer has connected its systems and the announcement has passed. Someone else arrives with a smaller budget and a use nobody planned for. They can find the terms, get the work they have a right to use and begin. Keeping that possibility alive would be worth paying for.
