Pseudonymous, not anonymous
Nothing in the math that turns a private key into an address, covered in Chapter 8, asks for a name. Generate a key, derive an address, receive coins — no identity field exists anywhere in that process for the protocol to check. In that narrow sense, Bitcoin really is permissionless and identity-free, and it's the sense most "Bitcoin is anonymous" claims are actually pointing at.
But that's only half the picture, and it's the half that makes the claim misleading rather than false. Every transaction ever made is permanently recorded on a ledger that anyone in the world can read, forever — the same public, independently auditable record Chapter 7 relies on to verify the 21 million cap, and the same one Chapter 9 traced UTXO by UTXO. Nothing about an address is private from observation; it's private only from attribution. Anyone can already see exactly how much a given address received, when, from where, and where it later sent funds — the only thing missing is a name attached to it.
That's a specific, third model, not a point on a line between "totally open" and "totally private." Physical cash leaves no ledger at all — a $20 bill changing hands is unrecorded and untraceable by design. A bank account is a ledger, but a private one, visible only to the bank, the account holder, and whoever the bank is legally compelled to show it to. Bitcoin is neither: a complete, permanent, public ledger that anyone can inspect without needing anyone's permission — full transparency of activity, combined with no built-in requirement to attach a real name to it. That combination is unusual enough that neither "like cash" nor "like a bank" describes it accurately.
An address isn't private from observation, only from attribution. The moment one point in a wallet's history gets a name attached, the ledger does the rest — backward as easily as forward.
What a KYC exchange already knows
Chapter 13 covered why a full exchange account requires identity verification in effectively every jurisdiction with meaningful trading volume — a legal requirement on the exchange, not a preference. The privacy consequence follows directly: the moment an exchange knows both your verified identity and the address you asked it to send bitcoin to, it has permanently joined those two facts in its own records, and by extension, so has the public ledger — because that withdrawal transaction, like every transaction, is visible to anyone who looks.
This is the same mechanism Chapter 19 discussed from the opposite angle — the reason blockchain-forensics firms and law enforcement have had real, well-documented success tracing and seizing Bitcoin tied to specific cases. There, it was framed as a reason Bitcoin is a worse choice than cash for large-scale crime. Here, the identical mechanism is the reason a KYC exchange withdrawal is the single most common way an ordinary, law-abiding holder's on-chain activity gets tied to their real name — not through any flaw in Bitcoin's design, but as the direct, unavoidable consequence of combining "everything is public" with "this one entity independently knows who you are."
Once that link exists, it doesn't just describe the withdrawal transaction itself. Because addresses in the same wallet are typically connected by earlier transactions — funds moving from one to another as part of ordinary use, exactly like the diagram above — the same identity link can, in practice, expose a meaningful slice of a wallet's history that predates the exchange relationship entirely, not merely what happens after it.
What actually improves privacy
None of this is unfixable, and several genuine, widely available mitigations exist — described here factually, without operational how-to or an endorsement of any specific tool or service, consistent with how this book treats every other category of product.
A new address for every incoming payment is already the default behavior in most modern non-custodial wallets (the category Chapter 12 covers), rather than something a user has to configure. It limits how much any single address's own history reveals on its own, though it doesn't fully sever the connection between a wallet's various addresses — which is exactly why the next point matters.
Avoiding unnecessary UTXO combination is the direct, practical version of the point Chapter 9 already made in detail: when a transaction spends multiple UTXOs together as inputs, that's a strong public signal they share an owner — commonly called the common-input-ownership heuristic, and it's one of the main tools blockchain-analysis firms actually use to cluster addresses into wallets. This chapter doesn't re-derive that mechanism; Chapter 9 already did, including the concrete fee-cost tradeoff of the opposite habit (consolidating small UTXOs to save on future fees, which is a privacy cost even when it's a fee-arithmetic win).
CoinJoin-style techniques exist specifically to break that same heuristic: several people's UTXOs get combined into one transaction with mixed outputs, so an outside observer can no longer confidently attribute any single output to any single participant the way an ordinary combined-input transaction allows. It's a real, meaningfully harder privacy tool than address hygiene alone, with real tradeoffs — coordination complexity, and in some jurisdictions, exchanges treating coins that have passed through a mixer with extra scrutiny. This book names the technique, not a specific implementation or service.
The Lightning Network, covered in full in Chapter 10, changes the picture differently: individual payments routed through open channels generally aren't written to Bitcoin's own public ledger the way an on-chain transaction is — only the channel's opening and closing transactions are. That's a genuine privacy improvement for the payments themselves, with the same real tradeoffs Chapter 10 already covered in depth (custodial Lightning wallets reintroduce a trusted third party, and this site's own tip-jar Lightning address is disclosed there as exactly that kind of custodial exception) — not repeated here.
A separate leak: the network layer, not the ledger
Everything above concerns what the ledger itself reveals — the recorded history of which address sent value to which. There's a second, independent leak vector that has nothing to do with the ledger's contents: the moment a wallet broadcasts a new transaction to the peer-to-peer network for the first time, whichever node happens to relay it first can potentially observe the IP address it came from. Well-resourced blockchain-surveillance firms are known to run many listening nodes across the network for exactly this purpose — linking the network origin of a broadcast to the transaction it carries, entirely apart from anything an address's own on-chain history discloses.
This matters because it's orthogonal to every mitigation already described. A wallet that never reuses an address, avoids unnecessary UTXO combination, and routes payments over Lightning can still leak an IP-to-transaction link at the moment of broadcast, if nothing addresses that specific layer. Two protocol-level responses exist, named here as techniques rather than an endorsement of any specific wallet's implementation: broadcasting over Tor, which many wallets support as a built-in option and which hides the true origin IP from the relaying node; and Dandelion++, a peer-to-peer relay design adopted by parts of the Bitcoin network that routes a new transaction through a short, randomized chain of relaying peers before it enters the normal broadcast pattern, making the true origin harder to pinpoint by the same kind of listening-node analysis.
Neither claim survives contact with how it actually works
The tabloid version — Bitcoin as the perfectly anonymous criminal's currency — doesn't survive Chapter 19's own facts: a permanent public ledger is precisely what let the FBI trace the Colonial Pipeline ransom and the DOJ trace funds tied to the Silk Road marketplace, years after the transactions occurred. But the opposite claim some proponents make — that Bitcoin is simply, fully private — doesn't survive this chapter's facts either: one KYC withdrawal, one careless public donation post, one data leak connecting an address to a name, and a meaningful slice of that wallet's history becomes exposed, forward and backward both.
"Permanent" is doing real work in that sentence, separate from "public." In August 2016, roughly 119,754 BTC was stolen from the Bitfinex exchange, then worth about $71 million, and moved through years of deliberate obfuscation on-chain. None of that trail ever disappeared — it just sat on the ledger, unattributed, until February 2022, when federal investigators combined blockchain analysis with conventional evidence to seize over 94,000 of those same coins, worth close to $4 billion at the time of seizure, and charge Ilya Lichtenstein and Heather Morgan, who later pleaded guilty to laundering them. Nothing about the ledger itself changed between 2016 and 2022 — the same public record that existed the day of the theft was still there, in full, six years later, waiting for the one identity-revealing thread that let investigators start walking it backward. A wiretap expires. A physical surveillance log gets discarded. An on-chain trail does not.
The honest answer sits in neither camp. Bitcoin is a public, permanent, pseudonymous ledger — a specific and genuinely unusual privacy model, not simply "more private" or "less private" than cash or a bank account on some single shared scale. It trades the bank's institutional secrecy for open verifiability, and it trades cash's total unrecordedness for a ledger anyone can audit without asking permission. Which of those tradeoffs matters more depends entirely on what someone actually needs — the same "argue the tradeoffs, not a verdict" pattern this book has used since Chapter 2.