Does a VPN make ChatGPT private? What it hides from OpenAI, and what it doesn't

The Morse teamPublished November 8, 20266 min read

No. As of November 2026, a VPN doesn't make ChatGPT private: it swaps your IP address, and the rough location that comes with it, for the VPN server's, and leaves your account, payment details, and every word you type with OpenAI. The location checks it was meant to beat don't all rest on your IP either. OpenAI also checks the country of your card, Anthropic goes by where you are "physically located," and Gemini can place you by the Home and Work addresses in your Google Account.

We read OpenAI's US privacy policy, its help pages on search, data controls, and unsupported countries, Anthropic's privacy policy and supported regions policy, Google's Gemini Apps Privacy Hub, and Mullvad VPN's no-logging policy, all on November 8, 2026. OpenAI deserves credit for saying plainly what a VPN does to ChatGPT's location estimate: it "may affect the approximate location inferred from your IP address" (OpenAI), and nothing on the page promises more.

What does a VPN change for ChatGPT?

The IP address OpenAI sees, and the "country, state, or city" ChatGPT estimates from it for local answers (OpenAI). Everything tied to your account stays as it was, and we compare a VPN with the methods that hide the account in anonymous ChatGPT. What your WiFi owner and internet provider see instead is in can the WiFi owner see your ChatGPT.

What does each AI company check that a VPN can't change?

Our table of the location and identity signals in each company's own pages, read November 8, 2026.

A VPN changes one row for the AI companies and adds a new party to the last one.

Does a VPN hide your location from Gemini?

Not reliably. Google's Gemini Apps Privacy Hub says "Location data is always collected if you use Gemini Apps," and that by default it comes from "the general area from your IP address or the Home or Work addresses in your Google Account" (Google). A VPN changes the first source, not the second, so a Gemini that knows your home address can still answer "What's the weather?" for your town.

Precise location gets past a VPN too. Gemini "may also receive precise location data from another Google service, like Google Maps, even if the precise location setting in the Google app is off." What Google stores with your activity is coarser, as we cover in is Gemini private. To change what Gemini falls back on, edit or remove the Home and Work addresses in your Google Account. For Claude, which collects your time zone setting and phone number alongside your IP, see how to use Claude anonymously.

Can you use ChatGPT or Claude through a VPN from an unsupported country?

Not within their rules. OpenAI's help page on unsupported countries, updated in late October 2026, says "Accessing or offering access to ChatGPT or the OpenAI API from an unsupported country may result in your account being blocked or suspended" (OpenAI). It adds a rule no VPN touches: "Using payment methods outside of our list of supported countries will result in you being blocked." A VPN can make your traffic look American, but it can't change the country your card was issued in.

Anthropic's rule turns on where you are. Its supported regions policy lists as unsupported any "Use by persons while physically located in an unsupported region," and it covers Claude.ai, Claude Code, and the API alike (Anthropic). By that wording, a VPN server in a supported country doesn't make your use supported.

Travelers hit the reverse problem. OpenAI says you may see an unsupported-country error "even when you are in a supported country or region," after your device connected "in an unsupported region and then return home," and its fix is to clear your browser's cache and cookies or try another browser.

Can a VPN get your ChatGPT account banned?

OpenAI publishes no rule about VPNs as such. Its privacy policy says it uses the area your IP points to "for security reasons," including "to protect your account by detecting unusual login activity" (OpenAI), so signing in from a new country through a VPN is the kind of change those checks watch for. The risk of a block comes from the rules above: an unsupported country, or a card from one. We found no OpenAI or Anthropic page that says using a VPN from a supported country breaks their terms.

What does the VPN company see?

Your real IP address and where your traffic goes, while you're connected. What it keeps afterwards is a promise, so read the policy before you pick one. Mullvad's is unusually specific. It lists "no logging of traffic," "no logging of DNS requests," "no logging of connections," and "no logging of IP addresses," and it checks your account's remaining time and connection count "in temporary memory only" (Mullvad). Its accounts are random numbers with no email or password, and it takes cash by post.

Mullvad also says where its own promise stops. If you pay by card or PayPal, the payment companies "log everything," so "your payment actions with these two methods are not anonymous." A VPN billed to your card knows who you are and where you connect from, which is the view your internet provider used to have.

How Morse keeps your identity away from the AI company

A VPN changes where your request seems to come from, and the AI company still files it under your account, your card, and your phone number. Morse sends the request under its own account instead of yours.

• Anonymous, for asking GPT, Claude, or Gemini as nobody in particular. Before your message reaches OpenAI, Anthropic, or Google, our proxy takes off your account, email, and IP address and sends it under Morse's own account, so the request carries none of your own account details. The provider sees the message without knowing it's you and may keep it for a limited time to check for abuse, under its own policy. Your words still arrive, so leave out your name, street, and employer.
• Private, the default, for when you'd rather no model host kept the chat at all. It runs open-weight models like DeepSeek, Kimi, and GLM, plus xAI's Grok, and Morse has zero data retention agreements with its inference, search, and other providers. Where a provider confirms zero retention on each reply, Morse checks it and refuses a reply that doesn't.
• E2EE, for a message you don't want any provider to read. Your browser checks the secure enclave's proof that it's genuine, encrypts the message on your device for that enclave alone, and our servers carry ciphertext. Third-party providers run the enclaves, not Morse, and the guarantee depends on the enclave hardware and attestation working as designed. E2EE comes with Pro, at $19 a month.

Your chats are stored in your browser, encrypted, rather than in an account on Morse's servers, and Morse doesn't train AI on your chats, sell your data, or show ads (privacy policy). Clearing the browser or deleting your account removes them for good, so keep a copy of anything you need. A VPN still earns its place on hotel or café WiFi, where it hides which services you use from the network.

Send a message from the Morse home page without an account, or sign up free with an emailed code.