Checking connection…
Security & Privacy

Can They Find Me If I Use a VPN?

What VPN providers really keep, what no-logs claims, open source and audits actually prove, and how payments can lead to identification.

Can they find me if I use a VPN? Laptop connected to a VPN with police officers in the background

As more and more countries pass new laws that continually restrict the privacy of communications and online activity, the need for meaningful privacy protection is growing ever more important.

The first and best solution that comes to everyone's mind is a VPN, and indeed, a good VPN service can effectively protect your privacy and anonymity.

However, choosing one is not as simple as buying a subscription through a sponsored link for a "big" and well-known provider. The market is shaped by multimillion-dollar advertising campaigns, affiliate commissions, and "review" websites of the "Best VPN 2026" type that have declared commercial relationships, or even common corporate ownership, with providers.

And as if that were not enough, the handful of business groups that can be counted on one hand, which according to approximate measurements¹ hold a market share exceeding 65%, have added new marketing elements to the selection process. Their sole purpose is to impress users, so that they buy a subscription believing they are "absolutely" safe and "invisible."

These elements are none other than open source client apps, independent audits, and, of course, the slogan "Zero Logging" or "NO-LOGS."

But let's start from the beginning and answer the most critical question on the minds of all VPN users.

Can They Find Me If I Use a VPN? The Short Answer

The short answer is NO, provided you are careful and well informed enough to make the right choices.

From the Exit IP to Identification: Which Data Can Be Linked?

When an online activity is linked to the public IP of a VPN server, the IP is usually the first available lead for an investigation. Of course, on its own it does not reveal who used the VPN. For an investigation to get from the exit IP to a specific person, there must be data linking the successive stages of the investigation.

There is no single mandatory order of requests for every case. Procedures depend on the country, the type of case, and the available data.

Generally, however, the most common procedure is:

  1. The website or system that received the traffic records the IP, the time, and possibly the port. This data only shows that the connection came from a specific public IP, not who the user behind it was.
  2. The IP leads to the entity that announces it or to the provider that hosts the server. The relevant records may show who rented the server or which provider the data center allocated it to.
  3. The network operator is asked whether it holds connection data. If it has records linking the exit IP and the specific time to a VPN account or session, an internal username or identifier may emerge.
  4. It is then checked whether this username or identifier can be linked to a customer account. For this to happen, a corresponding match is needed in one of the VPN provider's systems or records.

Consequently, each stage of the investigation requires the right data in order to move on to the next. If one link in this "chain" does not exist or was not retained, that path may end there.

However, things are not as simple as one might think. There are not only direct ways of reaching the end user, but indirect ones as well.

How a VPN User Can Be Exposed Through Logs

The Direct Way: What the "No-Logs" Policy Really Means

What does a zero-logging policy really mean, and how reliable can it be?

"No-logs" is not a technical term. It is simply a marketing slogan.

There is no common definition of no-logs. No standard, no regulator, and no law defines what a VPN provider is allowed to mean by this phrase. Every provider words it however suits them, and "logging" usually covers only what the provider itself considers logging. Typically that means browsing history, DNS queries, and/or traffic content. Almost all providers do exclude these, and in that respect the slogan is reliable and true. Everything else, however, and there is a lot of it, is left out.

There are at least six categories of data that a VPN can keep without violating its advertised "zero-logging policy," such as:

  • Connection metadata: when you connected, for how long, from which country or ISP, and how many MB you transferred. This data does not say "what you did," but it does say that you did something.
  • Account data: email, username, identifier, registration and renewal dates, logins to the user area, network connections, and subscription status.
  • Payment data: who paid, when, how much, and by what means.
  • App telemetry: crash reports, analytics, advertising IDs, device model, and OS version. Often, this data is also collected through third-party SDKs.
  • Support data: messages, attachments, and diagnostic files.
  • Website data: cookies, trackers, and analytics, which the policy often does not distinguish from other data.

And of course, there is a technical limitation that always applies. For the server to route your traffic, it must know your IP, at least for as long as the connection lasts. Also, for the provider to enforce the allowed limit of simultaneous connections per account, it must record somewhere how many connections are active at any given time. To deal with abuse or fraud, it must also keep some other data.

So some data always exists. Three things can make the difference:

  • Whether the data is kept only in memory or written to disk,
  • How long the data is retained, and
  • Whether it can be linked to your identity, which, in quite a few cases, it can.

That is why the question "does it keep logs?" makes no sense. The question to ask is which logs it keeps, for how long, where and how it keeps them, and whether and how the data can be linked to a person. And this is where providers' policies differ the most.

The Indirect Way: What 10 Well-Known Providers Actually Keep

No VPN provider states in its policy that it keeps browsing history, and that is probably true. It would, after all, completely defeat the purpose of the service it sells. One well-known VPN service states that it collects domains, but only in anonymized, aggregated form.

We mention this only to stress that very few users read a service's policies and terms of use and, most importantly, even fewer fully understand what they read and what it means for their privacy.

To help our readers understand exactly what we mean, we present a list of items drawn from the websites of 10 well-known providers that often appear in VPN review and comparison lists. The reading and analysis were carried out on October 25, 2026, and these providers are: NordVPN, Proton VPN, Mullvad, ExpressVPN, CyberGhost, Surfshark, PIA, IPVanish, Hotspot Shield, and TunnelBear.

We will not attribute the findings below to each provider, because this article does not aim to show whether a provider is "good" or "bad," but to help the reader understand what may be logged (browsing history aside, which we take for granted is NOT logged) and how that data can indirectly lead to the identity of a VPN user being revealed. Anyone who wishes can look up the findings below on each provider's website and verify them independently. That is why the items are listed in random order.

  • One of the providers mentioned above states that it keeps an indication of VPN use for the last 90 days and data from the last session for 15 minutes. For fraud checks, it keeps the first 8 and last 4 digits of the card and the billing IP. If a lawful request contains a user identifier, it says "whatever is available" may be disclosed.
  • One of the providers mentioned above states that it does not keep IPs permanently. However, its general policy provides for temporary logging to prevent service abuse and fraud, and says that, to verify the details of a real person, it may temporarily use their IP, email, or phone number. The telemetry of its apps is enabled by default.
  • One of the providers mentioned above states that card payments (e.g., Stripe, PayPal, Swish, and bank payments) are not anonymous because the processors keep records. Here the risk of correlation and identification lies in the payment method.
  • One of the providers mentioned above states that it keeps, per customer, the day of connection, the VPN location, the customer's ISP, and the total data volume in MB. Its partners may also receive the user's IP.
  • One of the providers mentioned above does not even state data retention periods. It says that connection events (Connection Attempt, Connection Success) and non-personal data such as the Google Advertising ID are recorded. It states that it allows analysis of user behavior (behavioral tracking) by third parties and does not specify which data concerns the website and which concerns the apps.
  • One of the providers mentioned above contradicts itself. In one section of its policy it rules out connection timestamps, while in another it states that it keeps a user ID or IP together with timestamps for up to 15 minutes after the end of the session. App telemetry passes through third parties.
  • One of the providers mentioned above simply states "zero user logs" but then contradicts itself by saying that it retains email and IP addresses without stating a retention period, saying only, vaguely, "as long as necessary." It also says that it may share data with companies in the group it belongs to.
  • One of the providers mentioned above states that it does not record browsing activity, but its apps store connection logs locally on the device. It says it can see them only if you send them. Payment data is kept for at least 7 years, and the policy allows data processors "anywhere in the world."
  • One of the providers mentioned above states that it records session duration and bandwidth per user. The device's IP is stored in encrypted form at the start of the session and deleted at its end. The free version includes advertising SDKs from eight networks, with advertising IDs and the device IP.
  • One of the providers mentioned above states that it keeps a monthly data total and an "active this month" indication, with no history. The website, however, uses Google Ads, Bing Ads, LinkedIn, and Hotjar, and the policy does not describe how to disable app reports.

If the "findings" above have anything in common, it is that almost none of them concern what you did through the VPN. They mainly concern who you are and how you behave online, and this can be inferred from your email, card, billing IP, advertising ID, payment processor, any connection timestamps, data volume, the identifiers the apps send, and so on.

A VPN may indeed have nothing to hand over about your traffic and the websites you visit, but if it has any of the data mentioned above and, most importantly, your payment details, it can link your identity to the VPN account if asked or compelled by the authorities. That is where your real digital traces lie: in the metadata the provider keeps, combined with your payments to it.

Follow the Money

The phrase "follow the money" describes one of the oldest and most effective methods used by law enforcement. When investigators cannot find the perpetrator through the act itself, they look at the transactions around it, because payments almost always leave a trace, are recorded, and are linked to real names, banks, and accounts. The same logic applies to a VPN. Even if the provider has no record of your traffic, your subscription is a transaction, and the transaction has a name, a card, a billing IP, and a timestamp.

Every provider that accepts payments keeps payment records. Otherwise it would not know whether you have paid, when your subscription expires, or who, for example, a refund request or chargeback dispute belongs to. This is clearly visible in the policies we presented. Quite a few providers keep payment and billing data for many years, because tax and accounting laws require it. When the provider itself does not store detailed data, the entity that settles the transaction does, such as Stripe, PayPal, Adyen, Braintree, Paddle, or the banks behind them. There you will find the cardholder's full name, the card number, the billing address, and the IP, and a lawful order from a judicial or prosecuting authority is enough to disclose them. The "we do not store card details" claim often applies to the provider, but not to the payment infrastructure it uses.

Consequently, the view that a VPN can "betray" you applies mainly to providers that keep some connection metadata, such as the day or time of connection, or the data volume. There, the payment acts as the connecting link: the account identified from the metadata is linked to the payment. When a provider keeps no such data at all, the investigators' job is much harder. At most, a court order can yield the IP from which the payment was made, but if that belongs, for example, to a second VPN, the chain stops.

In practice, we estimate that most identifications result from the user's own mistakes rather than from the VPN, provided that the VPN truly keeps no metadata at all.

There is no reliable research on how many VPN users have been identified. The cases we know of come from court rulings and from transparency reports that providers themselves publish, and there are certainly gaps in them. The fact that few cases have become known does not mean that there are really only a few. It is easy to see that, among the millions of VPN users worldwide, it cannot be that only 4-5 have been exposed. It is just that, for various reasons that are beyond the scope of this article, only these have become widely known.

The Role of Cryptocurrencies in VPN Payment Anonymity

Because paying for a VPN subscription by card indisputably means that your details are recorded and stored for many years, quite a few providers have begun to offer alternative payment methods, such as cryptocurrencies. Cryptocurrencies do offer a significantly higher level of anonymity, but not an absolute one.

For example, the king of cryptocurrencies, Bitcoin, is not anonymous but pseudonymous. That is, every transaction is public and permanent on the blockchain, and there are blockchain-analysis companies that specialize in linking wallet addresses to user identities. The link is usually made at the points where the cryptocurrency was acquired. For example, an exchange that requires KYC knows who bought the BTC and which address received it, so if the same address is later used for a VPN payment, the link is made automatically.

The Lightning network improves privacy considerably, because payments are made off-chain, but this too does not offer "absolute" anonymity, because your wallet provider (if it is custodial) knows what you are paying for. Also, most altcoins and stablecoins (Ethereum, USDT, etc.) are just as transparent as Bitcoin.

Monero is a different case because it is designed differently. Stealth addresses and ring signatures hide the sender, the recipient, and the amount, so that they do not appear on the blockchain. Its protection, however, is not absolute, because the point of acquisition and network metadata remain weak links. And when the payment also passes through a third-party crypto payment processor, that processor usually asks for an email address, records your IP, and may have its own KYC requirements, so the advantage is lost.

Nevertheless, payments via Lightning / Monero are considered very safe today, as long as the user does not make a mistake and the provider also has its own infrastructure for accepting payments in cryptocurrency, designed so that it keeps no data that is not actually needed to confirm the payment.

So, assuming that a provider keeps no activity history, and that the metadata it keeps to operate its service, as it has designed it, cannot be linked to a user identifier or to the user's payment, it is practically impossible for it to provide information that would identify the user.

Open Source in VPNs: What It Proves and What It Doesn't

Now that we have analyzed in depth the first marketing element, the advertised "zero-logging" policies, let's move on to the second marketing element, which is used to convince users that they are "absolutely safe."

The second element is none other than open source client apps. The argument sounds logical, and it catches on. Since the code is public, anyone can check what the app does, so the provider has nothing to hide. Open source has real value, but it proves far less than a VPN service may promise.

It is an indisputable fact that the open source code of a client app says nothing about what happens on the server. The client runs on your device. The logs, however, if they exist, are kept on the provider's servers, for example in user authentication systems and payment systems, and those remain closed. There is no way to see what is really running on a server and how it is configured. Auditing the client can tell you what your app sends, but not what the provider does with what it receives.

Also, the code you see is not necessarily the code you install. The app you download is usually a compiled file for your operating system, and the average user cannot easily check whether it was built from the published code. Open source presumes that someone actually reviews the code. For an ordinary user to do that, they would have to download the code from GitHub, set up the right development environment, build the app, and compare the result with the file they installed, also verifying its digital signature.

When was the last time you did anything like that? Most users have no idea where to start, even with ready-made instructions, and closed operating systems (iOS, Windows, macOS) make verification even harder, since the app stores also sign apps separately.

In practice, the "open" app remains, for most people, an act of faith in the provider's word. Open source does not mean a secure VPN.

So Does Open Source Have No Value at All?

We are not claiming that. It has value, but it is technically limited and certainly does not guarantee the "absolute anonymity" promised by advertising campaigns.

Open source simply lets someone see a very limited part of the overall service, and it is not an audit. It allows independent researchers to see whether the app contains third-party SDKs, what telemetry it sends, how it manages keys, how it handles the device's network connections, and so on. It is only an indication, and only about the app, not the whole network and the provider's infrastructure.

What would really matter? Independent audits that cover all servers and not only the apps, along with the software the servers actually run, databases, reproducible builds, and so on. And this brings us to the third marketing element, independent audits, which we turn to next.

Independent VPN Audits: What They Prove and What They Don't

As we said earlier about the NO-LOGS "slogan," the same applies, with even greater vagueness, to the third marketing element: the phrase "Independently Audited." It sounds like an official certification, but it certainly is not, because here we have the paradox of "made-to-order" certifications. For every independent audit carried out, an equal number of different "certifications" have been issued, each with its own audit scope.

That is, there is no common definition of "audit" or of what it must include. No standard, no regulator, and no law defines which checks must be carried out on a VPN provider.

Do you see the paradox here? Since all providers promise anonymity and ultimately sell exactly the same service, shouldn't there also be a common audit scope? In other words, shouldn't everyone be audited on exactly the same things?

And yet, each provider literally "orders" whatever audit it wants, and only what it specifies is audited. Nothing more. Of course, every audit report states its scope. But who reads those? Almost no one. And the result? The user simply sees the slogan "Independently Audited" and assumes everything is fine. This, however, is one of the industry's most misleading practices.

Let's look at it all in practical terms and as simply as possible. The word "audit" covers two very different things, and the average user has no idea what exactly they mean or how they may affect their privacy.

The first is the general security audit. Teams of experts, such as Cure53, X41 D-Sec, Assured, NCC Group, and Leviathan, usually examine the code and configuration of an app (and only very rarely an entire infrastructure) and look for vulnerabilities. These audits only tell you whether something is built correctly, not what is logged.

The second is the audit of the "no-logs policy." It is carried out by large firms such as Deloitte, based on the ISAE 3000 standard, which concerns assurance and not actual technical testing.

That is, it does not verify that "there are no logs." It simply checks whether the provider's statement matches its infrastructure. Well, someone might wonder, isn't that the same thing? Not exactly. To make this clearer for the average user, let's look at some real examples of "independent audits."

We reviewed recently published audit reports and announcements from the same providers mentioned above. Here, too, we do not name the provider behind each finding. Anyone who wishes can cross-check them on the providers' websites.

  • One of the providers presents an audit by one of the large auditing firms. It is an assurance audit (ISAE 3000, "reasonable assurance") explicitly described as "point in time," valid for a single specific date. The criteria were defined by the provider itself, and the auditor examined whether the description of the policy matched the servers' configuration, based on employee interviews and configuration files supplied by the provider. Out of scope were, among other things, access rights review, physical security, the streaming infrastructure, and the web/API, and use of the report is restricted to the provider and its customers.
  • One of the providers above presents an audit by another large firm, of the ISAE 3000 "Type I" variety. It examines the design and implementation of controls at a single point in time, not whether they worked properly over time. It covers the system of servers that run in RAM, not the entire company or its other systems (e.g., databases or payment records).
  • One of the providers above claims an audit consisting of employee interviews and inspection of the servers and technical logs, carried out over a few weeks. The number of servers examined is not stated, and the full report, according to its website, is available only after logging in to an account.
  • One of the providers above presents not one or two audits, but many, with different scopes: apps, infrastructure, web app, and a source code review of the payments and accounts API. Most of the audits concern code and configuration security and are not a statement that "there are no logs."
  • One of the providers above presents an audit based on three days of on-site work by two technical consultants. A random sample of production servers was examined, rather than the entire infrastructure. The auditors saw the systems as the provider's engineers showed them, not through independent analysis, and the criteria came from the provider's own public no-logs policy. Explicitly out of scope were CI/CD processes, source code review, and the payment and account management systems.
  • One of the providers above announces its third audit by a large firm, without naming the standard, the dates, or the methods. The publication describes the provider's design characteristics (RAM-only servers, regular reboots) rather than the audit procedures. It does not clarify whether the full report is publicly available, or where, as only a summary was found.
  • One of the providers above presents an audit via a press release. It states which claims were "verified" (no traffic, DNS requests, or user IPs are logged), but not the audit standard, the audit period, the audit methods, or the limitations. The auditor's report is not included, so the claims cannot be verified.
  • One of the providers above presents an audit by a cybersecurity firm, covering the source code and the partner platform of its group. The audit checked for logging of sensitive data and for the absence of activity and connection logs, using gray-box testing. The published page is a summary by the group itself rather than the auditor's report. It does not name the VPN service, and it characterizes the audit as "point in time."
  • One of the providers above presents the eleventh security audit by the same firm. The report explicitly excludes any audit of logging and data collection practices, as well as of whether the privacy statements hold true. Among other things, it found an unencrypted database containing account details and tokens in an Android app, and a data warehouse configured for public access.

Where Does All This Lead?

Audits differ greatly from one another in type (assurance or pentest), in time coverage, in who defined the criteria, in the sample, and in whether the report is public. In quite a few of them, the systems that are excluded are exactly the ones that hold your identity: accounts, accounting, payments, and web/API. An audit may confirm that the servers do not write logs, but it says nothing about whether, and where, your name can be linked to your account. This brings us back to "follow the money."

The Unified Business Model of Mainstream VPN Services

Many mainstream VPN services present customers with a single, unified experience. The user creates an account, provides an email address and a payment method, manages their subscription, and gets access to the apps and the VPN from the same account, within the same business ecosystem.

These services are often featured in comparison lists such as "Best VPN for 2026," in commercial or affiliate rankings, and through large advertising programs. Appearing in such lists, however, merely shows commercial reach. It does not prove by itself how a company handles data or whether it can ultimately link a subscriber to a specific VPN session.

In a unified model, the VPN provider holds all the data it has gathered about a user within a single corporate ecosystem (and possibly a single database). This can make it easier to link the various pieces of data, which can ultimately lead to the user's identity being revealed.

It is an indisputable fact that every provider maintains a database of its customers. And of everything we have mentioned so far, this is the most important point, and the one we have never seen any provider make public.

Namely, the structure of the database it maintains and which data tables it contains.

The more unified a VPN service's business model is, the more "unified" the information it holds about each customer.

A Few Words About K1VPN

K1VPN operates and provides its VPN services using a more "decentralized" administrative and business model.

Three different and independent companies, in three entirely different parts of the world and in three different jurisdictions, work together to deliver the VPN service to end users.

K1VPN LLC, based in Wyoming, United States, holds the license to use the K1VPN trademark, is the "seller" of the VPN services, and holds only limited subscription payment records, and nothing else.

Another company, in a different jurisdiction outside the US and outside agreements such as 5 Eyes, 14 Eyes, etc., is responsible for operating the VPN network and holds only the data that is absolutely necessary exclusively to provide the VPN service. It has no connection to, or access to, payment data or the records of K1VPN LLC.

Finally, a third company, on a different continent and in a different jurisdiction, owns the trademarks and certain intellectual property rights related to the service and the apps. It has access neither to payment data nor to VPN network data, and only oversees the quality of the services provided.

In addition, K1VPN, as the seller of the VPN services and the party responsible for managing payments and subscriptions, runs its own server for accepting cryptocurrency payments, which technically records nothing more than a successful payment linked to a randomly generated user account.

This "decentralized" model helps separate payments (follow the money) from use of the VPN network itself, and any lawful request for data must be routed through at least two different companies and jurisdictions.

This should in no way be taken as encouragement to use K1VPN for any purpose that violates our terms of use, on the "logic" that you will never be exposed.

Frequently Asked Questions

Can they find me if I use a VPN?

No, as long as your provider holds no data in its databases that could be linked to your payment and lead to a user being matched with a specific activity. You should carefully study your provider's privacy policy and which data it handles, and how.

Am I safe if my provider's client app is open source?

Open source software can only show whether the application is free of vulnerabilities and what data it sends to the provider. It cannot show you how the provider handles that data, so it is in no way a guarantee of security and anonymity.

If a provider has been independently audited, can it guarantee my anonymity?

You should carefully check the scope of the independent audit. If the no-logs policy has not actually been audited, then no other audit that has been carried out can guarantee your "absolute" anonymity. The most important thing to look at is the data held in the provider's databases and how it can be linked together.

If I pay by card and give my email, am I safe and anonymous?

It depends on the data the provider keeps in its databases, how long it is stored, and on what medium. Examine carefully which data is collected, how long it is kept, and where. Card payment records, even when the provider does not retain them, are kept for many years by banks and are linked to a person.

If I pay anonymously with cryptocurrency, am I safe and anonymous?

Cryptocurrency payments dramatically increase your level of anonymity, but they are not completely anonymous. For example, Bitcoin and other coins are not anonymous but pseudonymous. Today, Lightning and Monero payments offer the highest level of payment anonymity available. However, if your provider uses a third-party payment processor, part of your anonymity is lost. Choose a provider that operates its own cryptocurrency payment system, like K1VPN.

¹ The percentage is based on the number of Google Play Store downloads of the large, well-known VPN providers, as reported on various review sites and similar sources.

Disclaimer: This article is not intended to disparage any provider. K1VPN LLC adheres to the presumption of innocence, under which no provider is considered bad until proven otherwise. The investigation and final judgment rest with the reader. The providers named in this article were drawn from review sites that surfaced in random searches for the phrase "what is the best VPN."