← All posts

What your IP address actually reveals — and what it does not

6 September 2026 · 6 min read · Networking

Look up your own address on almost any site and you will be shown a city, a pair of coordinates and a map pin. It looks precise. Some of it is solid fact, some of it is an educated guess, and the two are almost never labelled differently.

Here is which is which.

Fact: the network that announces you

Every routable address on the internet sits inside a prefix that some autonomous system announces via BGP. This is not a database somebody compiled; it is the routing table that decides where your packets physically go. If it were wrong, you would have no connectivity.

From it you get, reliably:

Separately, the registry's own record — queryable over RDAP, the successor to WHOIS — gives the organisation the block is registered to, the exact range and the allocation type. Also fact, though it describes who holds the block, which is not always who is using it.

Estimate: the city on the map

There is no authoritative source for where an IP address physically is. No registry records it. Geolocation providers infer it, and they are working from registration data, network latency measurements, information ISPs volunteer, and user-reported corrections.

The consequences:

Anycast makes this stranger still. A single address like a public DNS resolver is announced from dozens of locations at once and answers from whichever is nearest. Ask two providers where it is and they will reasonably give different answers, because there is no single correct one.

The tell: when the sources disagree

Because geolocation and routing data are independent, comparing them is informative. When the geolocation provider says one country and the registry that allocated the block says another, something sits in between.

It could be a VPN or proxy. It could be an anycast deployment. It could be a block that was transferred between regions and never re-recorded. The disagreement is real evidence; the explanation is not something the data can settle on its own. Any tool claiming to have detected a VPN from this alone is overstating what it knows.

Reverse DNS proves nothing by itself

A PTR record maps an address back to a name, and it is often the most human-readable clue about an address — hostnames frequently encode the city, the datacenter or the connection type.

But whoever controls the address range writes the PTR, and they can write anything. The check that gives it weight is forward confirmation: resolve the name the PTR gives, and see whether it resolves back to the address you started from. If it does, the operator controls both directions and the name means something. If it does not, the name is a claim.

This is not academic. Many mail servers reject or penalise senders whose reverse DNS is missing or does not forward-confirm.

What is not revealed

Worth being clear, because the map pin implies more than it delivers. An IP address does not give anyone your name, your street address, your device, or your browsing history. Your ISP holds the subscriber record linking an address to an account at a point in time, and in most jurisdictions releases it only under legal process.

What the address does do is act as a stable identifier across every site you visit, which combined with a rough location and your ISP is enough to correlate your activity between unrelated services. That is the real privacy consideration, and it is not solved by the city being inaccurate.

If you want to change what it shows

The What's My IP module shows your public address and separates the two categories: routing and registry facts on one side, geolocation estimates on the other, with a note when the sources disagree. It also checks the published Tor exit list and the cloud providers' own range files.