The Age Check Appeared on Hotel Wi-Fi but Vanished on Mobile Data

The age-verification page appeared the moment my phone rejoined the hotel Wi-Fi. I had been viewing the same adult site on mobile data less than a minute earlier, using the same browser, account, and saved login. Nothing had changed except the network icon at the top of the screen. I refreshed, closed the tab, and opened the site again. The age prompt returned. When I switched Wi-Fi off, my library opened immediately on 5G.

I was staying in a hotel outside Birmingham before an early train.

The site held a video I had already purchased, and I wanted to download it for the journey. Mobile reception in the room was weak, so the transfer kept stalling. The hotel Wi-Fi was much faster.

Unfortunately, the faster connection was also the one that kept asking me to prove my age.

I assumed the browser had lost the earlier verification.

Then I switched back to mobile data and watched the prompt disappear again.

That was the clue.

The site was not reacting to my phone.

It was reacting to the route my phone used to reach it.

Wi-Fi and mobile data looked like different visitors

Since July 25, 2025, services that allow pornography in the UK have been required to use strong age checks before showing restricted content. Those checks can include facial age estimation, identity documents, banking information, credit-card checks, or confirmation from a mobile network operator. ()

From my side, the site saw the same adult with the same account.

From the site’s side, the two connections looked different.

On mobile data, the request travelled through my phone company’s network.

On hotel Wi-Fi, it left through an address shared by guests, staff devices, televisions, conference rooms, and anyone else using the property’s connection.

Web security systems can challenge traffic based on its public IP address, network owner, country, and reputation. Shared addresses with a history of unusual activity are more likely to attract extra checks. ()

That explained why the prompt could appear on one connection and not the other.

The hotel network had not discovered my age.

It had given the site a different reason to ask.

The hotel’s public address carried everyone’s history

I checked the public address on both connections.

The mobile network showed a large consumer carrier.

The hotel connection showed a business internet provider whose exit appeared to serve more than one property.

That distinction mattered more than the Wi-Fi signal strength.

A hotel guest may browse normally, while another device on the same public address sends automated requests, triggers rate limits, or repeatedly fails logins. The next guest inherits none of that behaviour personally, but the address can still inherit the reputation.

The age-verification page was only one possible response.

A site could also show a CAPTCHA, request another login, or refuse to trust a session that had begun somewhere else.

I had seen all three while travelling.

Other users have noticed the same strange fix: an age prompt appears on Wi-Fi, then disappears when the phone switches to mobile data. In one public discussion, the difference was traced to where the mobile carrier’s traffic actually left its network. ()

The practical lesson was simple.

“Wi-Fi” and “mobile data” were not privacy categories.

They were two network identities with different histories.

My established VPN replaced one crowded address with another

I already subscribed to a major VPN provider, so I connected it before trying the hotel Wi-Fi again.

The service had a long public history, mature applications, and a large server network. It was the obvious first choice.

Its automatic mode selected a nearby UK server.

The adult site loaded quickly.

Then the age prompt appeared.

I completed the facial-estimation check.

The result was accepted.

The site returned me to the library, displayed a CAPTCHA, and then asked for my age again.

I changed to a second UK server.

This time the library opened, but the download stopped after the hotel connection weakened. The VPN reconnected through another address, and the age prompt returned.

The provider’s strength was its ability to find a replacement route quickly.

For this task, that strength became the problem.

The verification began through one shared VPN address and continued through another. The site saw another abrupt change during a sensitive handoff.

I had escaped the hotel’s crowded public address only to arrive through a crowded VPN exit.

More servers meant more guessing

I tried Manchester.

Then London.

Then Edinburgh.

One server produced a CAPTCHA.

Another loaded the site but restarted the verification after I signed in.

A third opened the library and lost it when the Wi-Fi dropped for several seconds.

The server map gave me dozens of alternatives, but no way to know which address had the cleanest reputation or which one the app would preserve during the next interruption.

The comparison I had made before the trip now felt irrelevant.

I had chosen the provider partly because it offered so many locations.

Yet I did not need another country or another nearby city.

I needed one route that the site would accept and the hotel Wi-Fi could not easily break.

That was a smaller requirement.

It was also harder for the large server list to solve.

The smaller app started with the situation

OnlydogVPN was still installed from an earlier test.

It had fewer locations, a shorter public history, and fewer independent reviews than my established provider. Those limitations were why I had treated it as a backup.

But the app did not begin by asking me to select a country.

It offered a preset for private browsing on public Wi-Fi.

I selected it and connected.

Then I reopened the adult site.

The age prompt appeared once.

I completed the check.

The confirmation page closed.

My library opened.

No CAPTCHA followed.

No second age prompt appeared.

I selected the purchased video and started the download.

Five percent.

Fourteen.

Twenty-seven.

The result mattered before I knew the explanation: the same hotel Wi-Fi that had repeatedly broken the earlier session was now carrying the download.

The working route stayed together

Halfway through the transfer, the Wi-Fi signal dropped.

The progress bar stopped at 43 percent.

With the established provider, this was the point where the app had returned through a replacement address and the site had asked for another check.

The smaller app reconnected.

The download resumed from 43 percent.

The verified library remained open.

Its connection uses HTTP/3-based transport designed to recover when the underlying network path changes, rather than rebuilding every interruption as an entirely new journey. ()

The app also used obfuscation to make the protected traffic less obvious on the public network.

The practical effect required no protocol diagram.

The hotel connection faltered.

The protected route recovered.

The age result remained useful.

I did not have to choose another server, reopen the library, or place my face in front of the camera again.

The download finished on the network that had caused the problem

The progress bar reached 70 percent.

Then 86.

Then 100.

I switched the phone into airplane mode and opened the downloaded video.

It played.

That completed the task I had been trying to finish for almost an hour.

Mobile data had avoided the age prompt, but it could not hold a fast enough signal in the room.

The hotel Wi-Fi had enough bandwidth, but its shared identity kept creating friction.

The smaller app gave me the useful parts of both: the hotel connection’s speed without forcing the session to keep inheriting the hotel’s route.

That was the result the larger provider’s server map had not produced.

The counter showed what else had changed

After confirming the file worked offline, I returned to the smaller app.

Its blocked-request counter had increased while I moved through the verification page and library.

Advertising and tracking requests had been stopped around the session.

That had not solved the age prompt. The stable, less obvious route had already done that.

But it answered a second concern that appeared naturally after the download finished.

An age-verification flow already involves sensitive activity. I did not want unrelated advertising and tracking services receiving every background request generated around it.

The adult site still received what it needed to verify access and deliver the purchased file.

The surrounding session became quieter.

That gave me a reason to keep the app installed after I left the hotel.

The site did not know whether I had tapped Wi-Fi

I could not observe the site’s internal filtering rules or the exact score assigned to either network.

I could see the pattern.

Mobile data arrived through a consumer carrier route that the site accepted without another prompt.

Hotel Wi-Fi arrived through a shared address that triggered age verification.

The established VPN replaced that address with other heavily shared exits and changed routes when the connection weakened.

The smaller app carried one accepted session through the same unstable Wi-Fi until the download finished.

The site was not reacting to the symbol at the top of my phone.

It was reacting to the network identity beneath it.

That is why age verification can appear on Wi-Fi and disappear on mobile data, even when the device, account, and browser remain the same.

For this trip, mobile data was not inherently more private, and hotel Wi-Fi was not inherently more restrictive.

One route simply looked more trustworthy than the other.

The useful VPN was the one that made the faster network stop behaving like the suspicious one.