On-chain investigationStablecoinsCensorship

Bitcoin promised no one could stop a payment. Stablecoin issuers froze $6.4 billion

We went through ten years of on-chain records to count who freezes crypto today and what happens to the money they stop.

At a glance
Since 2016, stablecoin issuers have frozen $6.4 billion on the chains we read. Blockchains themselves froze, moved or rolled back $755 million. So 89% of the money stopped on-chain was stopped by an issuer, and Tether alone accounts for 97% of the issuer side. For 77% of frozen dollars we found no public reason. Freezes also tend to come late. TRON wallets frozen within a week of their first USDT had already sent out 75% of it.
89%
Share of frozen dollars frozen by stablecoin issuers
$6.4B
Stablecoins frozen by their issuers since 2016
97%
Share of that frozen by Tether
77%
Frozen dollars with no reason we could find
Key findingsEach line has its own share link
  1. 89%

    Stablecoin issuers froze $6.4 billion on-chain since 2016. Blockchains froze, moved or rolled back $755 million.

  2. 97%

    Tether made 97% of the issuer freezes by value, most of them on TRON.

  3. 77%

    For 77% of frozen dollars we found no published reason. About 9% was tied to US sanctions.

  4. 75%

    TRON wallets frozen within a week of their first USDT had already sent out 75% of it. On Ethereum it was 41%.

  5. 83%

    Of $1.4 billion in frozen USDT that Tether destroyed, 83% went back out of its Treasury, to the cent, most of it into Bitfinex.

01 · Two walletsTether froze $344 million in one afternoon

If you hold USDT, the company that issues it can freeze your coins. On 23 April this year, Tether did that to two wallets holding $344 million between them. Nobody had to stop a blockchain. The coins just stopped moving.

Both wallets were on TRON, the chain where Tether does most of its freezing. A day later the US Treasury's sanctions office, OFAC, added both to its list under its entry for Iran's central bank. Tether said it had worked with OFAC and US law enforcement. They are the two largest single freezes we found.

Two wallets, one afternoonThe 10 largest single freezes of a USDT or USDC wallet, value at the moment of the freeze
TRON USDT, 23 Apr 2026$213MTRON USDT, 23 Apr 2026$131MEthereum USDT, 13 Jan 2022$100MEthereum USDT, 20 Nov 2023$87MTRON USDT, 14 Jul 2026$86MTRON USDT, 10 Feb 2026$84MTRON USDT, 12 Jun 2026$72MEthereum USDT, 20 Nov 2023$65MEthereum USDT, 13 Jan 2022$60MTRON USDT, 15 Jul 2024$55M
Red: the two TRON wallets Tether froze on 23 April 2026. Value is the USDT or USDC in the wallet when it was frozen, at $1 a coin.

Bitcoin's white paper set out to remove exactly this kind of control. It describes cash sent straight from one person to another, with no bank in the middle. Stablecoins don't work that way. They are dollar tokens issued by companies, known as stablecoin issuers, and the issuers kept the power to stop them.

We counted every freeze we could find on-chain over ten years. Some were by stablecoin issuers, some by the chains themselves. We put a dollar value on each one as of the day it happened. Stablecoin issuers froze $6.4 billion. Blockchains froze, moved or rolled back $755 million. That puts nine in ten dollars stopped on-chain in the hands of issuers, and Tether alone stopped 97% of the issuer side.

02 · The switchCan stablecoins be frozen?

Yes, and it takes one transaction. A stablecoin is a token that tracks a currency, most often the US dollar. Tether issues USDT, the largest. Circle issues USDC, and Paxos issues USDP and PAXG, a gold token. Each token is a smart contract that keeps its own record of who holds what, on TRON, Ethereum and many other chains.

These contracts come with a blacklist. Once Tether adds a wallet to it, the wallet's USDT cannot be sent. Tether can later take the wallet off the list, or burn the frozen balance and pay the same amount to someone else from its Treasury. Paxos has the same switches, and Circle can freeze and unfreeze USDC. The chain keeps running the whole time.

Blockchains can stop money too, but it takes more people. On a proof-of-stake chain such as Sui or BNB Chain, validators are the computers that run the network. They can agree to freeze a wallet, halt the chain or roll it back to undo a hack. Most of them have to go along. A softer form needs no agreement at all. Block builders and miners pick which waiting transactions go into each block. One of them can leave a transaction out of its own block, and the transaction waits for someone else's.

Two places where a payment can be stoppedWhat a stablecoin issuer can do, and what a blockchain can do
The stablecoin issuerTether, Circle and Paxos keep a listof blocked wallets inside their own token.One call adds a wallet to the list.Its USDT or USDC cannot move after that.The issuer can later unfreeze it.Tether and Paxos can also burn thebalance and send the same amount elsewhere.Works on any chain the token lives on.The chain itself keeps running.The blockchainValidators can agree to freeze anaddress, halt the chain, or roll it backto undo a hack.Block builders and miners can chooseto leave some transactions out oftheir blocks.Hits every coin on that chain.Freezes, halts and rollbacks need most ofthe network to agree.
Left: the freeze and unfreeze switches built into USDT, USDC and the Paxos tokens, and the burn switch in USDT and the Paxos tokens. Right: freezes, halts and rollbacks by validators, and transactions left out by block builders and miners.

03 · The countWho freezes crypto: mostly stablecoin issuers

For each dollar a chain stopped, a stablecoin issuer stopped more than $8. Most of the issuer side sits on just two chains, TRON and Ethereum.

Nine in ten dollars, frozen by companiesEvery $100 frozen or moved on-chain since 2016, by who did it
$89 frozen by stablecoin issuers$6.4 billion in all$11 frozen or moved by blockchains$755 million in all
Stablecoin issuers: freezes of USDT, USDC and the other Tether, Circle, Paxos and TrueUSD tokens, valued at the moment of the freeze. Blockchains: freezes, halts, forced moves and rollbacks with a dollar figure.

Issuers froze more almost every year, and hit a record $2.4 billion this year. The chains acted in bursts instead, most of them after big hacks.

Issuer freezes hit a record in 2026Dollars frozen or moved each year, by stablecoin issuers and by blockchains
$0$1.0B$2.0B$3.0B20162017201820192020202120222023202420252026 *$2.4BThe DAO forkBNB Chain freezeStablecoin issuersBlockchains* 2026 to 16 October
Value at the moment of each action. 2026 runs to 16 October. Blockchain years count only actions with a dollar figure.

04 · TetherTether does almost all the freezing

Tether froze 97% of the issuer total. Its USDT on TRON alone makes up 64%. Circle and Paxos, the next two, froze far less.

USDT on TRON leads by farStablecoin value frozen since 2016, by token and chain
USDT on TRON$4.1BUSDT on Ethereum$2.1BUSDC on Ethereum$149MPAXG (gold) on Ethereum$20MUSDC on Base$14MUSDP on Ethereum$4.7MBUSD on Ethereum$2.2MUSDC on Arbitrum$2.0MTUSD on Ethereum$1.8MXAUT (gold) on Ethereum$1.0M
Value at the moment of each freeze. Repeat freezes of a wallet that is still frozen count once. Tokens and chains under $1 million are left out.

Counted by number of freezes, Tether made 69% of them. Many of Circle's freezes copy its Ethereum list onto newer chains like Base and Arbitrum, where those wallets hold almost nothing. We went through every USDT freeze on Ethereum and TRON in an earlier piece.

05 · The reasonsWhy does crypto money get frozen?

Tether and Circle seldom say why they froze a wallet. We matched each freeze to the OFAC list as it stood that day and to published police and court cases. Then we checked our address labels. For 77% of frozen dollars we found no reason at all.

Most frozen money has no public reasonDollars frozen by stablecoin issuers, by the reason we could find
No reason we could find77%Police or court request7.1%OFAC listed it within a week6.1%Scam or fraud3.5%Hack or theft2.6%Already on the OFAC list2.1%Tied to a sanctioned group0.6%Laundering or terror funding0.5%Mixer0.1%
Each freeze gets one reason, in this order: on the OFAC list that day, listed in the 7 days after, a published case, then our labels. One-word labels such as "scam" alone do not count.

Sanctions explain only a small part. Wallets already on the OFAC list make up 16% of all freezes but only 2.1% of the dollars. Most of them were empty when frozen. The bigger sanctions share comes from freezes like the opening one, where OFAC listed the wallets in the days after the freeze. All the sanctions cases together come to about 9%.

When OFAC does list a wallet, Tether now freezes it fast, most often the same day. Since 2024, 76% of its freezes of listed wallets came inside 7 days of the listing. It has frozen every TRON wallet on the list at some point.

Token and periodFreezes of listed walletsMedian days after listingWithin 7 days
Tether USDT, before December 20235167 before0%
Tether USDT, December 20231583960%
Tether USDT, 2024 onward395076%
Circle USDC on Ethereum, before 2024151197%
Circle USDC on Ethereum, 2024 onward65097%

Our labels point the same way. 17% of frozen wallets carry a crime or sanctions label, and they hold 13% of the frozen dollars. 77% carry nothing that says who owns them. A few hundred belong to ordinary businesses, like exchange deposit wallets, with no crime label at all.

Most frozen wallets are unknownFrozen addresses and frozen dollars, by what our labels say the address is
Share of frozen walletsShare of frozen dollarsCrime or sanctions label17%13%Ordinary business2%1%Contract or named account4%5%Nothing says who it is77%82%
Each wallet counted once per chain. Crime or sanctions: an OFAC listing or a hack, scam, laundering, mixer or sanctioned-exchange label. Ordinary business: exchange, payment or gambling labels with no crime label.

Even that split is too kind to the labels. Wallets still frozen carry a label of some kind far more often than ones that were later let go, so some labels were likely added after the freeze. Some freezes also end. 11% were later reversed, after a median 96 days. In late March and early April, Circle and Tether froze 529 wallets, then lifted all of those freezes by late May.

06 · The speedDoes freezing catch the money in time?

A freeze only helps if the money is still there. For each wallet's first freeze, we checked how much of the USDT it got had already left. Wallets frozen within a week of their first USDT had already sent out 41% of it on Ethereum and 75% on TRON. Wait half a year, and little is left.

Late freezes find little leftShare of the USDT already sent out when the freeze came, by time from the first USDT to the freeze
0%25%50%75%100%41%75%Within 7 days72%85%8 to 30 days72%96%31 to 180 days88%99%Over 180 daysEthereumTRONDelay from a wallet's first USDT to its freeze
Each wallet's first freeze, counting the USDT it received before that freeze. Sent out means it left the wallet before the freeze; it does not show who ended up with it.

Even same-day freezes leak. Freeze a wallet the day its first USDT arrives, or the day after, and 37% of the money has still left on Ethereum. On TRON it is 54%. And few freezes are that fast. Only 23% of frozen wallets were caught in their first week, and on TRON that share has fallen since 2023.

Freezes are not getting fasterShare of frozen wallets caught within 7 days of their first USDT, by year of the freeze
0%10%20%30%40%18%27%25%28%22%15%32%32%23%19%EthereumTRON20222023202420252026 ** 2026 to 16 October
Years before 2022 had too few frozen wallets on TRON to compare. 2026 runs to 16 October.

These dollar shares lean on a few big wallets that pass money through. The 10 largest frozen TRON wallets received 65% of all USDT that reached frozen TRON wallets. Wallet by wallet, the two chains split. Half of frozen Ethereum wallets sent nothing before their freeze, while the median TRON wallet had already sent on 92%.

Hacks show both sides. When the thief left the coins in place, a fast freeze worked. Tether and Circle froze $119 million from the KuCoin, Poly Network and Multichain thieves, each within about a day. Thieves who kept moving had little frozen. Less than 1% of the $47.6 million in stablecoins taken on Ethereum in the Bitget hack was frozen. In the Ledger wallet drain, Tether froze 14% of the stolen USDT.

HackChainHours from first funds to freezeFrozenMost held at onceStablecoins stolen
KuCoin hackEthereum14$19.8M$19.8Mnot known
Yearn yDAI exploitEthereum17$1.7M$1.7Mnot known
Poly Network hackEthereum4$33.4M$33.4Mnot known
BNB Chain bridge exploitEthereum5$5.1M$15.2Mnot known
Multichain exploitEthereum25$65.8M$65.8Mnot known
Remitano hackEthereum3$1.4M$1.4Mnot known
Remitano hackTRON3$0.5M$0.5Mnot known
Bybit hackEthereum15$0.6M$0.6Mnot known
Bybit hackTRON8$4.6M$5.5Mnot known
Bitget hackEthereum8$0.3M$0.9M$47.6M
Bitget hackTRON74under $0.1M$0.2Mnot known
Ledger/CryptoBilis drainTRON2$10.0M$10.0M$70.2M

07 · AfterWhat happens to frozen USDT?

Most frozen USDT stays frozen. Of the USDT Tether froze in wallets on Ethereum and TRON, 67% is still frozen. Tether destroyed 22% and released 12% back to the holders. Releases came far faster on TRON, a median 49 days after the freeze. On Ethereum the median wait was over a year.

Frozen wallets also keep getting paid. A frozen USDT wallet cannot send, but it can still receive. $330 million arrived in frozen wallets after their freeze, and more than half of it is still stuck.

Most frozen USDT is still frozenWhat happened to the USDT Tether froze in wallets, by chain
Ethereum22%38%40%TRON7%14%79%Released to the holderDestroyed by TetherStill frozen
Value at the moment of the freeze. Destroyed counts what Tether burned, which can include money that arrived after the freeze. Zero address and USDT contract balances are left out.

When Tether destroys a frozen balance, the same amount usually goes back out of its Treasury. Tether's Treasury is the wallet it issues new USDT from. For 83% of what Tether destroyed, the Treasury paid out the exact same amount, a median 5 minutes from the burn. Most of those payouts, 94%, went to addresses that pass money straight into the crypto exchange Bitfinex. On-chain the trail ends inside the exchange, so we cannot see whose account was credited. To follow money like this yourself, ask the Bitquery MCP server.

Burned, then paid out againWhere the money goes after Tether destroys a frozen balance in a wallet
$1.38Bdestroyed from frozenwallets on Ethereumand TRON$1.15B83% paid out again bythe Tether Treasury, sameamount, within minutes$1.08B94% of that went toaddresses that pass itstraight into BitfinexThe rest$55M matches a round payout within an hour.$133M (10%) has no matching payout within 180 days.$33M went back to Poly Network.$29M went to Tether payout hubs.$41M matched only round payouts long after, which chance also produces. Not counted.
A match is a Treasury payout of an equal amount near the time of the destroy, to the millionth of a USDT for most of the value. No record on-chain links the two.

The biggest match fits the US Secret Service seizure made public in June 2025. It covers 7 Ethereum wallets Tether had frozen 19 months before. Tether burned the coins in 6 of them, and 27 minutes later the Treasury paid out the sum of all seven, about $225 million, to one address. That address sent it on to a Bitfinex hot wallet 12 seconds later. Tether burned the seventh wallet's coins 3 hours after the payout.

08 · The chainsWhen blockchains freeze, halt or roll back

The other side of the count is small. Blockchains step in rarely, and mostly after big hacks. The biggest case was BNB Chain in 2022. After a bridge hack, its validators halted the chain and blocked the hacker's wallet, trapping $289 million in BNB.

Chain actions are rare and mostly follow hacksValue frozen, moved or rolled back by blockchains themselves
BNB Smart Chain 2022$289MSui 2025$162MCronos 2026$111MArbitrum One 2026$72MEthereum 2016$45MJuno 2022$36MGnosis Chain 2025$9.4MHive (Steem fork) 2020$9.2MSteem 2020$5.7MKiiChain 2026$5.3MSteem 2020$3.2MFogo 2026$3.0MCosmos Hub 2026$2.1MBitcoin Cash 2019$1.5MMANTRA 2026$0.2M
15 of the 34 chain-level actions we found since 2016 count toward the total. BNB Chain and Arbitrum are measured from our data; the rest use reported amounts. KiiChain and Fogo announced their actions, but we could not confirm they were carried out. A freeze and the later move of the same funds count once.

Other chains did the same. In 2025, Sui's validators froze $162 million stolen from Cetus, a DEX on Sui, and a vote later sent it back. This April, Arbitrum's Security Council, a group that can push emergency changes, stepped in after a hack. It moved $71.6 million in ETH out of the Kelp DAO exploiter's wallet. In August, Cronos rolled back its chain to undo the Tectonic exploit.

Some of that money goes back. Of the $755 million, $207 million went back to the people it was taken from, and $121 million partly did. For most of the rest, including BNB Chain's, we could not find out.

09 · The blocksIs Bitcoin censorship resistant?

Bitcoin's promise rests on a different layer: who gets to put payments into blocks. That kind of censorship leaves no freeze behind. Ethereum is where it has been tested most. There, block builders put together most blocks, and some of them leave out transactions that touch sanctioned wallets. Those transactions can still get into someone else's block. We tested each builder, quarter by quarter, for blocks with fewer sanctioned transactions than chance would give.

Up to 64% of Ethereum blocks came from builders that left out sanctioned transactionsShare of Ethereum blocks built by builders that leave out sanctioned transactions, by quarter
0%20%40%60%80%2022202320242025202664%Builders that leave them outAfter the Tornado Cash delisting: no firm answer
A builder counts when its blocks hold clearly fewer sanctioned transactions than chance would give. About 15% of blocks could not be tied to a builder, so shares are floors. Grey: after the Tornado Cash delisting, too few such transactions remain for a firm answer.

For nearly two years, builders that leave them out made about half of all Ethereum blocks, and 64% at the peak. Since the US lifted its Tornado Cash sanctions, so few such transactions are left that the test gives no firm answer.

On Bitcoin, there is little to test. Since late 2018, only about one Bitcoin block in 590 held a payment to or from a wallet on the OFAC list.

Bitcoin has little to censorBitcoin blocks holding a payment to or from an address on the OFAC list, by year
010020030092018 *92019152020462021104202229320231492024432025462026 ** 2018 from 28 November; 2026 to 16 October
714 of 420,023 blocks since 28 November 2018. An address counts from the day OFAC listed it. Only 112 of the 533 Bitcoin addresses on the list ever sent or received bitcoin while listed.

That is too few to say whether any mining pool leaves them out on purpose. One pool, F2Pool, said it had filtered such payments. In December 2024 it mined none of the 11 such blocks it would normally have found. So is Bitcoin censorship resistant? Mostly, yes. A bitcoin payment cannot be frozen, and a payment one pool leaves out can still go into the next block another pool finds.

That is the split this count shows. Blockchains rarely stop anyone. The issuers of the stablecoins on top of them freeze thousands of wallets a year, mostly without saying why. If you hold USDT or USDC, the company that issues it decides whether your coins can move.

10 · How sureHow sure are these numbers?

The headline holds when we change the rules. Counted 5 other ways, the issuers' share stays between 84% and 95%. It falls to 76% only in one case. That is if we count the aUSD stablecoin a bug minted on Acala, a chain in the Polkadot network, at face value. We treat it as worth nothing.

The same answer, counted seven waysStablecoin issuers' share of all frozen dollars under different counting rules
70%80%90%100%Our count89%Add the EOS 2021 and THORChain cases84%BNB Chain at its public $429M figure88%Leave out USDT later given back88%Chains: only what we measured ourselves95%Add company freezes our data cannot see90%Count Acala's bug-minted aUSD at face value76%
Each dot is the whole count redone with one rule changed. The first dot is the count used in this story.

Some freezes are missing. Tether and others have reported $676 million in freezes our data cannot see. The largest is a Tether freeze for Turkey. Adding them raises the issuer share to 90%. Our Tether totals match our earlier census of every USDT freeze. They also agree with the 2025 address counts of BlockSec, a security firm that tracks USDT freezes.

11 · MethodHow we did this

We read every freeze, unfreeze and burn in ten tokens on Ethereum: seven stablecoins, two gold tokens and Circle's cirBTC. We also read USDT on TRON, and USDC on Base, Arbitrum, Optimism and Polygon. The records come from Bitquery's on-chain data for each chain, starting with Ethereum. Each freeze is valued at the wallet's balance at that moment, priced that day. A wallet frozen again while still frozen counts once.

Chain actions come from the chains' own posts, and from our data for BNB Chain and Arbitrum. The block tests compare each builder's or pool's blocks with what chance would give. The full transfer histories are also sold as files on the Bitquery Data Store, for Ethereum and TRON.

12 · EvidenceTransactions and wallets cited

WhatWallet or transaction
Larger wallet frozen on 23 April 2026TNiq9A…ZH81
Its freezeebe670f1…75fe
Smaller wallet frozen the same minuteTTiDLW…jSr9
Its freeze295cd606…b09a
USDT on TRONTR7NHq…Lj6t
USDT on Ethereum0xdac1…1ec7
Tether Treasury on Ethereum0x5754…b949
One of the 7 destroys, 18 June 20250x923657…172d
Treasury payout of 225,364,963.88 USDT0xeae979…7b88
Address that received it0x62f5…0d2f
Its transfer to Bitfinex0xdcd8a3…afc1
Bitfinex hot wallet0x7713…35ec

FAQ

Can Tether freeze my USDT?

Yes. USDT has a blacklist built in. Once Tether adds a wallet, its USDT cannot be sent. Tether has frozen $6.2 billion this way on Ethereum and TRON so far.

Can Circle freeze USDC?

Yes. USDC has the same kind of switch. Circle froze $166 million in USDC on the chains we read, often copying the same list of wallets onto each new chain.

Is crypto censorship resistant?

Coins such as bitcoin and ether are hard to stop. Stablecoins are not. Of the money stopped on-chain since 2016, 89% was frozen by stablecoin issuers, and 11% by blockchains.

Is Bitcoin censorship resistant?

Mostly. A bitcoin payment cannot be frozen. Only about one Bitcoin block in 590 held a payment tied to a wallet on the OFAC list. That is too few to tell whether any mining pool keeps them out.

Why does Tether freeze wallets?

It rarely says. For 78% of the dollars Tether froze we found no published reason. About 9% was tied to US sanctions and 7% to published police or court cases.

What happens to frozen USDT?

Most of it stays frozen. 67% of the USDT Tether froze in wallets is still frozen. When Tether burns a frozen balance, it usually pays out an equal amount again within minutes.

Limits on these figures

Issuer freezes are counted on Ethereum, TRON, Base, Arbitrum, Optimism and Polygon. USDT on other chains and freezes on Solana are not in the totals.

Each freeze is valued at the wallet's balance at the moment of the freeze. Money that arrives later is not added, so the totals undercount what ends up stuck.

Most chain-level dollar figures come from the chains' own reports. We measured BNB Chain and Arbitrum ourselves. Actions with no published amount add nothing to the total.

Reasons come from public sources and our labels. A freeze with no reason we could find may still have one that was never published.

Sent out before the freeze means the money left the wallet. It does not show who ended up with it.

A destroy-to-payout match is an equal amount at nearly the same time. It is strong evidence, not a record that links the two.

Block tests show which builders and pools include fewer sanctioned transactions than chance would give. They cannot show why.

Run it yourself

Ask who froze what, in plain English

Every token freeze in this story comes from Bitquery's on-chain data. The Bitquery MCP server puts that data behind an AI assistant, so you can ask when a wallet was frozen, or where a reissued balance went, without writing the query yourself. The full transfer histories are also sold as files on the Bitquery Data Store.

Check whether a wallet was ever frozenList the freezes of a stablecoin on a dateFollow a reissued balance to its next walletCompare a wallet with the OFAC list
Explore Bitquery MCP →Figures measured from on-chain data from 2016 to 16 October 2026, checked on 18 October 2026.
Legal disclaimer

This article is provided for informational and educational purposes only and reflects analysis of publicly available on-chain data as of the dates indicated. It does not constitute legal, financial, compliance, or investment advice, and nothing in it is a recommendation to buy, sell, or hold any token or asset.

The findings describe freezes, destroys, halts and rollbacks observed on-chain from 2016 to 16 October 2026. Wallets are described by their behavior on the chain and by our labels. None of them is attributed to a named person.

Tether, Circle, Paxos, TrueUSD, Bitfinex, F2Pool and the blockchains named appear because their tokens, wallets or networks appear in the records described. That says nothing about the conduct of those companies beyond what the records show.

Reported by Gaurav Agarwal for Bitquery Research, with AI tools; every figure was checked against the raw data.

Nothing herein should be relied upon as a definitive determination of fact. Readers should conduct their own independent verification before taking any action. The authors and publisher accept no liability for any loss or damage arising from reliance on this material. All trademarks and company names are the property of their respective owners.