Best No-Logs VPN: A Privacy Checklist

Learn how to verify no-logs claims, minimize the account and payment data you leave behind, and stay safer on public Wi-Fi.

What makes a good no-logs VPN? Look for a service that explains which data is stored, what must be retained, who can access it, and when it is deleted. A claim that it “doesn’t log browsing activity” doesn’t tell you whether connection times or your entry IP can be linked to your account. Compare traffic, connection, account, and diagnostic data separately.

What data to check in a no-logs policy

Servers need to process data to route a connection, but processing it doesn’t mean keeping it long term. A privacy policy should distinguish temporary state needed for an active connection, fields written to logs, and records that can still be queried later. Check whether fields can be linked: a timestamp alone may not identify a user, but combined with an entry IP, account identifier, and exit records, it could help reconstruct a connection.

Compare no-logs policies by data type
Data typeFields to checkQuestions worth asking
Browsing activityDestination domains, destination IPs, DNS queriesDoes “we don’t log browsing activity” also exclude destination addresses and DNS queries?
Connection metadataEntry IP, connection start and end times, exit route, usageDoes “for operational purposes only” mean aggregate statistics or individual connection records that can be looked up?
Account and transactionsAccount identifiers, order numbers, payment recordsWhich fields are kept by the service or payment provider? Can they be linked to connection records?
Diagnostics and supportCrash reports, app logs, support ticket attachmentsAre diagnostics uploaded by default? After opting out, is similar data still sent?

Distinguish “usage totals accumulated per account” from “browsing details retained.” The former may be used to deduct data from a plan and doesn’t, by itself, reveal which sites you visited. The latter may include destination addresses and timestamps that can be used to trace activity. Don’t automatically accept a claim of “anonymous analytics”: first check whether raw records are kept before aggregation and whether they contain identifiers that can track the same device over time.

How to verify privacy claims

Start with the current privacy policy and note its effective date and scope. The website, apps, and network infrastructure may collect different data. Website analytics terms don’t replace the network policy, and an app’s support for an encryption protocol doesn’t tell you what the servers retain.

  • ✅ Check the fields. Confirm the policy explains how it handles entry IPs, connection times, destination addresses, and DNS queries.
  • ✅ Check retention periods. For data that is stored, find out when it’s normally deleted, how backups are handled, and what exceptions apply.
  • ✅ Check who receives the data. Find out whether third parties handle hosting, diagnostics, customer support, or payments, and which fields they can access.
  • ✅ Check the evidence. If the provider publishes an audit report, review when it was conducted, which systems it covered, and any exceptions. An audit limited to the client app doesn’t provide the same evidence about network infrastructure.
  • ❌ Don’t use an exit IP test to infer what the provider retains. A test only shows where that request exited; it can’t verify whether servers write logs.

If you can’t find an answer, ask a question that can be answered with specific fields: “After a connection ends, do you still store entry IPs and connection start and end times that can be linked to an account? If so, for how long?” If the written reply conflicts with the published policy, ask which clauses apply rather than assuming the more favorable interpretation.

Bottom line: Rule out services that don’t explain key data fields or retention periods, then compare routes and prices. Being unable to verify a claim doesn’t prove that records exist, but it also doesn’t prove that nothing is retained.

How to minimize account and payment data

Check what’s required first, then how account recovery works. Not requiring an email address means you don’t have to link an additional contact identifier to your account, but find out in advance how to recover access if you lose your credentials. Use a unique password and keep recovery information somewhere you control. If your client imports routes using a subscription link, treat that link like a credential: anyone who gets it may be able to view your subscription configuration while it remains valid.

Payments create a separate data trail. Payment providers may keep transaction records under their own policies, while the service may use an order number to activate service, reconcile accounts, or process refunds. Before paying, check what each party receives and whether orders are retained for accounting purposes after you request account deletion. “No browsing logs” doesn’t mean “no payment records,” and don’t provide false information to try to reduce the data collected.

How to check your connection on public Wi-Fi

A no-logs policy describes what the provider retains. Public Wi-Fi risks depend on where your device connects, which traffic goes through the tunnel, and what happens if the connection drops. Even with a working tunnel, signing in to a website still identifies your account to that site, and changing your exit IP doesn’t remove existing browser cookies.

  1. Check the network name first. Confirm the hotspot name with the venue instead of connecting to one that merely looks similar. If a captive portal is required, complete only the necessary steps. Don’t bypass a certificate warning just to continue.
  2. Then check split tunneling. In the client, see which apps or domains connect directly. Seeing the expected exit IP in your browser doesn’t mean background sync tools use the same route. If a work app needs to use the tunnel, check its rules before using it.
  3. Check the DNS route. Run a DNS test when you’re not handling sensitive tasks, then compare the results with your client settings. A browser’s secure DNS may use a separate resolver, so seeing a different provider doesn’t necessarily mean there’s a leak. The key question is whether queries bypass the route you intended.
  4. Test what happens if the connection drops. If the client offers disconnect protection, enable it and use a page with no sensitive content to check for unexpected direct connections during a network switch. Manually disconnecting and losing the connection unexpectedly may trigger different behavior, so testing only one doesn’t confirm the other is protected.

If you can’t confirm how split tunneling or disconnect behavior works, pause syncing and sensitive tasks until you’re back on a trusted network. On company devices, follow your organization’s network and data-handling rules; don’t replace managed settings with a personal configuration.

Start Free