Every wallet scam that has ever worked — every drainer, every fake support desk, every “verification” page — relies on a single precondition: the victim not knowing which of the wallet’s strings does what.
Not weak passwords. Not slow reflexes. The gap is vocabulary. A wallet hands you a seed phrase, keeps a private key, exposes a public key, and prints an address — four things that sound like the same thing to a newcomer and do completely different jobs. Scammers don’t defeat the cryptography. They exploit the fact that you might paste the wrong string to the wrong window.
This article walks the chain from top to bottom — what each string is, what leaking it actually costs, and the attack pattern attached to each layer of confusion. The crypto is simple. The consequences of mixing them up are not.
The One-Way Chain
Everything in a wallet derives from everything above it, in one direction:
seed phrase → private key → public key → wallet addressThe seed phrase (BIP-39, 12 or 24 words — defined by the Bitcoin Improvement Proposal of the same name) is the master secret. From it — via the hierarchical derivation of BIP-32 — the wallet deterministically derives one or more private keys, each controlling exactly one address. Each private key produces a public key through elliptic curve math (secp256k1) that is instant forward and infeasible backward. The public key is then hashed down — Keccak-256, last 20 bytes on Ethereum — into the 0x... string you give people when you want to receive funds.
Four layers. Each step downward throws away information, which is why none of the arrows reverse. You can’t compute the public key from an address, can’t compute the private key from a public key, and can’t recover a seed phrase from anything below it. That irreversibility is the security model.
Now the practical tier list, which is the whole article compressed:
| String | Moves money? | Safe to share? | What leaking it costs |
|---|---|---|---|
| Seed phrase | Yes — all accounts | Never, no exceptions | Everything, permanently |
| Private key | Yes — one address | Never, no exceptions | That address, permanently |
| Public key | No | Yes, by design | Nothing directly |
| Address | No — receive only | Yes, by design | Targeting: profiling, dust, poisoning |
Every row transition is a scam. Read on.
Layer 1: The Seed Phrase and Private Key — the Things That Move Money
If you remember one sentence from this article: the seed phrase and the private key are the only strings in your wallet that can move money, and that is why nothing legitimate will ever ask you for them.
There is no verification flow, no wallet “sync,” no KYC process, no airdrop eligibility check, no customer support script, and no blockchain “validation” that requires your seed phrase or private key. None. The phrase exists to be written down once, offline, and typed back into wallet software you chose to trust during recovery. Any other context where it appears — a website form field, a support chat, a “secure verification” modal, a DM — is the harvesting step of a scam.
The industry has spent a decade making this simple, and scammers have spent the same decade making it sound complicated. Their favorite framing is procedural: the phrase is needed to “verify ownership,” “connect your wallet to the network,” or “register your address.” All of these translate to one action: you hand the attacker your wallet. A well-run drainer site will then empty the address within minutes, often faster than you can finish reading the confirmation page — and since the transfer is a legitimate signature from the legitimate key, no chain, exchange, or explorer will treat it as theft.
If you used an online generator or a chatbot to create your phrase, that’s a separate and equally fatal variant of the same mistake — we cover the full burned-phrase protocol in why you should never generate a seed phrase online.
Layer 2: The Public Key — the String You’ll Almost Never Touch
Here’s the awkward truth about the public key: for all the fear newer users have about “leaking” it, you will almost never handle it directly, and on Ethereum it isn’t even private in practice. Every signature you broadcast mathematically reveals enough to recover your public key — which is why Etherscan can display it for any address that has ever sent a transaction. On Bitcoin, a standard address hides the public key until the first spend from it reveals it.
The public key’s job is verification. When your wallet signs a transaction with the private key, the network checks the signature against the public key — confirming the sender controls the matching key without ever seeing the key itself. It’s elegant, and for you as a user it is entirely invisible plumbing.
So why does the public key matter in a scam article? Because of the request. “Please share your public key so we can verify your wallet” is one of the most reliable scam scripts in circulation, and it works precisely because the term sits in the victim’s blind spot: technical enough to sound like a real procedure, unfamiliar enough that the victim doesn’t know which of their strings it refers to. The attacker isn’t hoping you find the public key. They’re hoping you don’t find it — dig around, get frustrated, and paste the one long secret-looking thing you do have: the seed phrase.
The tell is that no legitimate consumer flow needs your public key from you, ever. Wallets derive it internally. Contracts verify signatures on-chain. If a website, support agent, or “airdrop checker” asks, the request itself is the attack.
Layer 3: The Address — Public by Design, Exploited by Design
The address is what the whole stack exists to produce: a short, shareable, receive-only identifier. Giving it out is normal — payroll, airdrop claims, transfers between your own wallets. Nobody can compute backward from it, and nobody can sign anything with it. Sharing an address is safe.
Trusting an address is not. This is the distinction attackers run on, and two attack families live entirely inside it:
Address poisoning. Because addresses are designed to be public and public addresses end up in your transaction history, an attacker generates thousands of addresses that look like yours — same first and last characters — and sends you micro-transactions so a lookalike lands in your history. Months later, when you’re sending a large transfer and copy an address “from a previous transaction” instead of from your address book, you paste the attacker’s. The chain executed your instruction perfectly. Our address poisoning explainer covers the defensive habits: address books, first-and-last character checks, and why test transfers are cheap insurance.
Dust and fake tokens. Your address being visible means attackers can push junk to it — unsolicited tokens designed so that interacting with them (trying to sell, swap, or “claim” their apparent value) routes you into a malicious approval. The unsolicited airdrop tokens guide covers the rule: tokens you didn’t ask for are bait, not windfall.
The Cross-Layer Confusion: Signing Is Not Sending
One more boundary scammers exploit, and it sits between the layers: the difference between sharing a secret and using a secret.
When your wallet signs a transaction, the private key never leaves the device — the signature is proof the key approved a message, mathematically unrecoverable from the output. Signing is how the system works. But the safety of a signature depends entirely on what’s being signed, and this is where people who’ve internalized “never share your key” still get drained:
- A token approval is a signature. Sign an unlimited one and you’ve delegated spending rights — see the infinite approval explainer.
- A permit2 message is a signature. Its whole innovation is moving the on-chain cost after your approval — the permit2 exploit breakdown shows what attackers do with that.
- A raw
eth_signrequest is a signature — of an arbitrary transaction the site can then broadcast. The eth_sign phishing case study is the canonical example. - Blind signing is the wallet-level enabler: hardware wallets displaying hashed gibberish because the payload isn’t decoded, training users to confirm what they cannot read.
So hold both rules at once, because each covers the other’s blind spot: never share the key, and never sign what you can’t read. Scammers build campaigns against people who know only one of the two — the cautious person who won’t share anything but signs blindly, and the careful reader of payloads who still pastes a phrase into a “verification” form.
The Self-Test
Run any request you receive through this filter, in order:
- Are they asking for an address? Can be legitimate — receiving funds, airdrop eligibility, payroll. Verify you’re on the real site before pasting it, then it’s fine.
- Are they asking for a public key? Almost certainly a scam in consumer contexts. No wallet, dapp, or support flow you’ll legitimately encounter needs it from you.
- Are they asking for a private key or seed phrase? Scam. Zero exceptions, no matter the stated reason, urgency, or authority. Verification, sync, migration, “blockchain registration,” inheritance unlock — the reason changes, the answer doesn’t.
- Are they asking you to sign something? Signing can be legitimate — it’s how you interact with any contract. So the question becomes: did you initiate this, on a site you verified, and can your wallet display the decoded payload? If any answer is no, treat it as a signature scam until proven otherwise.
The filter works because it isn’t about judging intentions — it’s about which string does what. That part never changes, which is what makes it the one piece of crypto knowledge that can’t be socially engineered around.
If You Already Pasted the Wrong String
The order of operations matters when seconds do:
- If you shared a seed phrase or private key — assume the wallet is already being drained. Create a fresh wallet (on a device you trust, with a phrase generated on-device), move anything remaining from an old transaction only if you know its origin, and treat every address under the old phrase as permanently burned. Then follow the first 24 hours runbook.
- If you signed something you shouldn’t have — the damage may be an approval, not a transfer yet. Revoke it (how to use approval revocation tools safely) before the attacker pulls funds, and move high-value assets to a clean address if the approval can’t be cleanly revoked.
- If you sent funds to a lookalike address — it may not be recoverable, but fast reporting can freeze funds at an exchange hop: the how to report a crypto scam walkthrough covers IC3, exchange freeze requests, and assembling the package investigators actually use.
In all three cases, expect the second wave immediately: recovery scammers monitor public complaints and contact fresh victims offering to “trace” the funds for an upfront fee. Anyone who contacts you first is running that play.
The Difference That Protects You
Strip away the cryptography and this article is about three nouns. The seed phrase and private key are the money. The public key is plumbing. The address is a mailbox slot.
The scams aren’t sophisticated — they’re auditions for your confusion. “Verify your wallet” auditions for the person who can’t name the public key. “Confirm this transaction” auditions for the person who signs without reading. “Use the address from your history” auditions for the person who forgot public doesn’t mean trusted. Every one of them loses the audition the moment you can say which string is which — and that knowledge takes ten minutes, costs nothing, and never expires.
That’s a better deal than the alternative. Just ask anyone who learned the difference one transaction too late.
Frequently Asked Questions
Is it safe to share my public key?
The public key itself is designed to be public — on Ethereum it is mathematically recoverable from any signature you've ever broadcast, and block explorers display it openly. The real danger is not leaking the public key; it is not knowing which string is which. When a scammer asks you to 'share your public key for verification,' the request is a fishing trip: they are hoping the victim who doesn't know the difference pastes the seed phrase or private key instead. No legitimate service ever needs your public key for anything you'd do as a normal user.
Can someone steal my crypto with just my wallet address?
No. The address is a receive-only identifier derived from your public key — it cannot sign anything, and no amount of computation will reverse it back to your keys. What attackers do with your address is different: they profile your holdings on-chain to target you, send dust and fake tokens to bait you toward malicious contracts, or run address poisoning so a lookalike address lands in your history hoping you'll copy-paste it during a real transfer. Safe to share, dangerous to trust.
Why do scammers ask for a 'public key' during wallet verification?
Because the request sounds technical enough to be plausible and harmless enough that an unsure victim will comply. The script works on people who know the phrase 'public key' exists but not which of their strings it refers to. In documented support-scam patterns, the victim digs around their wallet, finds the only long secret-looking thing they have — the seed phrase — and sends that. The defense is a flat rule: addresses can be shared; public keys are almost never legitimately requested; private keys and seed phrases are never requested by anyone for any reason.
Is signing a transaction the same as sending my private key to the website?
No — that confusion cuts both ways and both directions get exploited. A signature is proof your private key approved a message; it is generated on your device and does not reveal the key itself. That's the safe direction. The unsafe direction: people who know 'signing is safe' get careless with what they sign. A signature can carry an authorization (a token approval, a permit2 spend, or in the eth_sign case a raw transaction), and blind signing hides the payload from you. Signing is only as safe as the thing being signed — read our signature scam explainer before you click confirm on anything you didn't initiate.
What's the difference between a seed phrase and a private key?
The seed phrase is the master secret: 12 or 24 words from which a wallet deterministically derives all your private keys, across every account the wallet manages. A private key controls exactly one address. Leaking either one is total compromise, which is why they sit in the same tier of the security model — the difference is scope, not sensitivity.