A misunderstanding has followed peer to peer since the Napster years: the idea that "no central server" automatically means "no one can see you." It's the opposite. In a network of peers your device talks directly to everyone else's, and to do that it exposes something very concrete: your IP address. Decentralizing, on its own, doesn't make you anonymous. Often it makes you more visible.
The core misconception: distributed isn't private
In a centralized service your data passes through an intermediary: that party knows everything about you, but there's a wall between you and other users. In a P2P network that wall is gone. Whoever takes part in the same exchange can, in principle, see who else is taking part. Trust doesn't disappear when you remove the center: it shifts onto other elements — the protocol, the cryptography, the discipline with which you use the tool. If you don't think about it, no one does.
Case 1 — BitTorrent: 148 million IPs from a single laptop
In 2010 a group of INRIA researchers presented a study at a USENIX workshop with a telling title: "Spying the World from your Laptop". From a single machine, over 103 days, they collected roughly 148 million IP addresses tied to BitTorrent downloads, and managed to identify who injected content into the network in 70% of cases.
The point isn't "BitTorrent is insecure": it's that in a swarm your IP is, by construction, public to every other peer. An entire industry grew out of that visibility: firms that monitor swarms, collect addresses and send settlement demands to users (the infamous copyright letters). No sophisticated hacking — just the open nature of the network turned against those who use it without protection.
In a network of peers, the most valuable information often isn't the content: it's the metadata. Who, with whom, when, from where.
Case 2 — Bitcoin and Silk Road: pseudonymous isn't anonymous
Bitcoin is perhaps the most misread example. Many used it thinking it was "anonymous money." In reality it's pseudonymous: every transaction is written forever onto a public ledger anyone can read. Addresses don't carry your name, but they're persistent labels — and the moment a single one of those labels is linked to your real identity, the whole connected history becomes readable backwards.
Silk Road is the textbook lesson. The online black market ran on Tor and paid in bitcoin, yet its founder Ross Ulbricht was arrested on October 1, 2013. He was caught by identity-management mistakes (old forum posts traceable to him) combined with the fact that the transaction ledger was, and remains, fully public. The analytical discipline that studies these flows — blockchain analysis — is today a mature trade, used by companies and law enforcement alike.
Case 3 — Monero and the $625,000 bounty
Precisely because Bitcoin's transparency is a privacy problem, networks were built from the start to be untraceable. Monero is the best known: ring signatures, stealth addresses and hidden amounts make it very hard to reconstruct who pays whom.
How well it works is told by a concrete fact: in 2020 the U.S. tax agency (IRS) offered up to $625,000 to anyone who could "break" Monero's privacy and trace its transactions. Around the same time the firm CipherTrace announced it had developed Monero-tracing tools for the Department of Homeland Security. When a state puts a bounty on a network's anonymity, that's the clearest sign the anonymity — designed well — is real.
Case 4 — Tor and I2P: adding anonymity at the transport layer
If the problem is your IP traveling in the clear, the answer is a layer that hides it. Networks like Tor and I2P route traffic through multiple nodes (onion routing) so that no single hop knows both origin and destination together. They're not magic — they have known limits, from exit nodes to traffic-correlation attacks — but they move privacy from "hoping no one is watching" to "making watching expensive." It's the same principle as P2P: don't trust a single point.
Case 5 — Communications: encrypting content isn't enough
Modern messaging almost always encrypts content end-to-end. But content is only half the problem: the metadata remains — who writes to whom, when, how often — and it often says more than the words. P2P projects like Briar, Session or Cwtch exist precisely to remove the intermediary servers and minimize this metadata, accepting in exchange the classic costs of peer to peer: peer discovery, NAT, availability when devices are offline.
How privacy is designed in a network of peers
There's no "privacy" switch. There's a series of choices, each with a trade-off:
Hiding the transport (Tor, I2P, VPN) protects your IP but adds latency. Breaking linkability (Monero's techniques, CoinJoin on Bitcoin) protects your financial history but requires dedicated tools. Minimizing metadata (serverless P2P messaging) protects your relationships but complicates the user experience. Good privacy comes from combining these layers according to what you actually want to protect — not from the illusion that decentralization offers them for free.
Decentralizing is a necessary condition for digital freedom, but not a sufficient one for privacy. The first removes the master; the second has to be built.
The takeaway
The thread running through these cases is always the same. BitTorrent shows the network exposes you; Silk Road shows the pseudonym falls if you get one detail wrong; Monero shows anonymity can be engineered — enough to earn a bounty; Tor and P2P chats show privacy is added, layer over layer. Peer to peer gives us autonomy back by removing intermediaries, but with autonomy comes responsibility: in a network without a center, privacy is something you choose and maintain yourself. Not a free side effect of the architecture.