Most crypto theft needs a hacked exchange, a leaked seed phrase, or a signature you shouldn't have approved. Address Poisoning needs none of that — just one copy-paste from the wrong line in your own transaction history. On 20 December 2025, one trader reportedly sent 49,999,950 USDT to a scammer this way, minutes after sending a careful 50 USDT test to the correct address.

The short answer
Address Poisoning is a crypto scam where an attacker plants a look-alike wallet address in your transaction history, usually through a zero-value or tiny transfer, and waits for you to copy it by mistake the next time you pay. Your keys are never stolen — you send the money yourself, to the wrong address, and the transfer can't be reversed. Protecting against Address Poisoning comes down to a single rule above everything else: never copy an address from your own history.

Why the attack runs almost entirely on bots

What makes Address Poisoning cheap enough to attempt millions of times is that it's fully automated. Bots scan networks such as Ethereum, BNB Chain and Tron looking for wallets that move USDT regularly, and they note who those wallets typically pay. The attacker then generates a vanity address matching the first and last few characters of your usual recipient — a modern GPU can find a four-character match at each end within minutes. A transaction plants that look-alike in your history, and the attacker simply waits. Sometimes the payoff takes days; in one February 2026 case flagged by security firm Cyvers, only 32 minutes passed between the poisoning and the actual loss. Once funds arrive at the fake address, they're typically swapped and scattered across multiple wallets almost immediately to make tracing harder.

Why the zero-value transfer trick is the most common method

This is the method that surprises people most, because it can make it look as though you were the one who paid the attacker. USDT and most EVM tokens follow the ERC-20 standard, which includes a transferFrom function meant to let an approved third party move tokens on someone's behalf. In many widely deployed token contracts, a transfer of zero tokens passes the approval check automatically, since zero can never exceed a zero allowance. That means anyone can trigger a "transfer" of 0 USDT from your address to theirs, with no money actually moving and no alert raised on your end. Your wallet or a block explorer may still list it as an outgoing payment from you, which is exactly what makes the fake address look like someone you've legitimately paid before. Some explorers hide zero-value transfers by default; a lot of wallet apps still don't. Two related variants work the same trap differently: a fake token contract literally named "USDT" that mimics the exact amount you just sent, and a dust transfer that sends a tiny real amount from the look-alike address to look like ordinary incoming money. All three forms of Address Poisoning leave the same result — a convincing fake sitting right next to the real address in your history.

Why USDT specifically draws this kind of attention

Attackers go where the money concentrates, and a lot of it sits in USDT. Large single transfers are common — a treasury move or an OTC settlement can send millions in one transaction, so a single mistake can cover the cost of millions of failed attempts elsewhere. Repeat recipients make the pattern easy to copy, since exchange deposits, supplier invoices and payroll-style payouts tend to go to the same handful of addresses again and again. Attempts themselves are cheap, often costing mere cents on busy networks, which lets attackers fire off enormous volumes and need only a tiny fraction to succeed. And because a confirmed blockchain transfer is final, there's no chargeback mechanism and no bank to call once funds are gone.

Why the reported scale of this problem deserves attention

Public reporting on Address Poisoning is striking even allowing for how differently each source measures it. One security firm reportedly flagged over 65 million poisoning attempts across roughly a 14-month window according to industry trackers, and a separate academic study counted close to 270 million attempts across two major chains over a two-year period, aimed at around 17 million potential victims, with tens of millions of dollars confirmed lost. These figures shouldn't be added together since different researchers count different things, but the pattern they describe is consistent: the overwhelming majority of attempts fail outright, and the rare successes pay enough to keep the automation running indefinitely.

Why spotting a fake wallet address is harder than people expect

An EVM-style address runs to 42 characters, and almost nobody reads the whole thing. Most people check the start and the end, and many wallets even shorten the middle to dots as a convenience. Attackers count on exactly this habit, building look-alikes that copy precisely the characters a typical user is likely to glance at. Shortened, a real address and a poisoned one can look identical even though the middle 30-plus characters are completely different. The warning signs worth knowing: a zero-value or tiny transaction from a sender you don't recognise, two near-identical addresses sitting in your history, a new entry appearing right after a genuine payment, a token labeled "USDT" that you never actually held, or a wallet warning about a new or similar address that you click past without reading.

Eight habits that actually stop Address Poisoning
Never copy an address from your transaction history. Get it from the recipient's own invoice, their exchange deposit page, or a QR code instead — history is the one place an attacker can write to, and this single habit change matters more than any other.
Save verified addresses in an address book after checking the full string once, then select saved contacts for every future payment rather than re-copying anything.
Turn on withdrawal allowlists wherever your exchange offers them, ideally with a waiting period applied before any newly added address becomes active.
Compare the whole address in chunks of four to six characters — start, middle and end — reading it carefully on the screen where you actually confirm the send.
Don't let a clean test transfer relax you. In the December 2025 case, the test payment went perfectly to the correct address, and the attacker struck before the main transfer went out — always re-verify the address again after any test, not just before it.
Filter out spam and zero-value transfers using the settings most wallets and explorers already offer, which clears out much of the noise attackers depend on.
Use a wallet with built-in look-alike detection. Since mid-2026, several major wallets compare every pasted address against ones you've used before and block the send with a warning if the start and end match but the middle doesn't.
Slow down on large transfers. Confirm the address through a second channel such as a call, or have a colleague double-check it. Address poisoning is fundamentally a speed attack, and a deliberate pause beats it every time.
A two-minute pre-send check worth building into habit

Before sending anything that matters, get the address from its original source rather than your history, confirm it's the right network (ERC-20, BEP-20, TRC-20 or similar), read the full address in chunks against that source, send a small test amount first, check your history for any new suspicious entries that appeared after the test, re-verify the address one final time, then send the remainder and confirm on a block explorer that the receiving address matches exactly.

Why businesses face a larger version of the same risk

Companies are attractive targets precisely because they pay the same suppliers repeatedly, and finance staff are often moving quickly under deadline pressure. The same copy-from-history habit that catches an individual catches a payout team too, just with much larger numbers attached. A maintained, approved payee list — paid only from stored addresses, never from chain history or a forwarded message — closes much of this exposure on its own. Requiring two approvers for any new or changed payout address, confirmed by a phone call to a number already on file rather than one just received, adds a meaningful second layer. Locking down payout automation with scoped API keys and two-factor authentication prevents a single leaked credential from being able to redirect funds. And a checkout or invoice that shows the exact deposit address plus a QR code means a customer never has to copy anything from a past transaction at all.

What to do if you've already sent USDT to a poisoned address

The honest answer is that recovery is unlikely, since blockchain transfers can't be reversed and attackers typically move funds within minutes. Acting quickly still gives you a small chance and protects you from the second wave of scams that often follow. Save the transaction hash, both addresses, the time and the amount as evidence, and avoid sending anything further from that wallet in a panic. Report the incident to your exchange immediately if one was involved, since compliance teams can sometimes flag the receiving address. File a report with your local cybercrime authority as well — Tether has frozen USDT at specific addresses before at the request of law enforcement, though that outcome needs a formal report and fast action, with no guarantee either way. And ignore anyone offering a paid "recovery service" afterward; these consistently target people who have just lost funds and rarely deliver anything.

FAQ on Address Poisoning

What is address poisoning? It's a scam where an attacker plants a look-alike wallet address in your transaction history, usually through a zero-value or tiny transfer. If you later copy that address instead of the real one and send USDT, the funds go to the attacker, and the transfer generally can't be reversed.

How does an address poisoning scam actually steal USDT? Bots watch wallets that move USDT regularly, generate an address matching the first and last characters of a usual recipient, and plant it in your history. When you copy the fake address instead of the real one, you send the funds yourself — no private key or seed phrase is ever needed.

Is my wallet hacked if I see a poisoned transaction? No. A poisoned entry only means an attacker wrote a fake record to the public blockchain — your keys and funds stay safe as long as you don't copy and send to the look-alike address.

Can I get my USDT back after an address poisoning attack? Usually not, since blockchain transfers are final and attackers move funds within minutes. Report it immediately to your exchange and local cybercrime authority, and never pay a "recovery service" — these are commonly scams targeting people who've just lost funds.

Does address poisoning happen on networks other than Ethereum? Yes. It mostly hits account-based networks where anyone can send a token to any address — Ethereum, BNB Chain, Tron and other EVM-compatible chains all see it. The defence stays identical everywhere: never copy an address from your history.

What's the single most effective way to prevent address poisoning? Never copy an address from your transaction history. Save verified recipients in an address book, use exchange withdrawal allowlists, compare the full address before sending, and use a wallet with look-alike detection built in.

Final thoughts

Address Poisoning doesn't exploit any flaw in the blockchain itself — it exploits a habit. Copying an address from history is quick and almost always harmless, which is exactly why it's so dangerous on the one day it isn't. Break that single habit, save addresses you've actually verified, read the middle of every address you're unsure about, and slow down whenever the amount is large. If you handle USDT payouts for a living, build the same discipline into your process so that one rushed afternoon can't cost a year of margin.