Blockaid detected the exploit in progress and put the face value of newly minted SAND at approximately $49 billion across more than 400 transactions.
PeckShield counted 14.9 billion SAND directed to two attacker addresses. One later analysis put the full mint at 329.24 trillion tokens across 703 events over five hours.
The attacker’s actual proceeds were roughly 80 ETH, about $675,000.
Why the Numbers Diverge So Wildly
Face value is a multiplication, not a theft. Take the number of tokens minted and multiply by the market price of the legitimate token.
That calculation assumes the tokens could be sold at that price. They could not. $49 billion vastly exceeds the available market liquidity, and SAND’s entire legitimate supply is 3 billion tokens on Ethereum.
The mint exceeded the real supply by orders of magnitude, which is precisely why it was worthless.
The extraction figure is the one that describes what happened. BlockWatchdog documented roughly 14.75 million SAND drained from the Ethereum OFT Adapter in under 60 seconds, converted to approximately 79.74 ETH.
| Measure | Figure | What it means |
| Face value minted | about $49 billion | Token count times market price, unrealizable |
| Tokens minted (Blockaid tally) | 14.9 billion to two addresses | First wave visible on-chain |
| Tokens minted (full count) | about 329.24 trillion over 703 events | Across roughly five hours |
| Actually extracted | about 14.75 million SAND | Drained from the Ethereum OFT Adapter |
| Proceeds | about 79.74 ETH, roughly $675,000 | The real loss |
| Project impact statement | under 0.01% of total SAND supply | Refers to legitimate supply affected |
What Actually Broke
SAND on Base is deployed as a LayerZero v2 Omnichain Fungible Token, which uses a delegate address on the destination chain holding administrative rights over endpoint configuration.
Blockaid reported that the attacker hijacked those LayerZero delegate permissions through the `approveAndCall` function, gaining the ability to mint on Base without corresponding collateral locked on Ethereum.
This is the same root cause behind earlier landmark bridge failures. Ronin, Wormhole and Nomad all involved an attacker gaining control of an administrative function that should never have carried unilateral minting power.
The category has not changed in four years. Only the specific permission model has.
What Limited the Damage
This is the detail worth sitting with, because it was not a security control.
In this bridge design, the Ethereum adapter holds only as many tokens as users have previously bridged across. On the night of August 21, that adapter held a limited amount of SAND.
Once the attacker drained those reserves, no further backed SAND existed to extract, regardless of how many unbacked tokens continued to be minted.
So the ceiling on the theft was set by how little liquidity had been bridged, not by anything the protocol did. A more successful bridge with deeper reserves would have lost more.
That is an uncomfortable observation and it should temper any reading of this as a containment success.
What the Team Did Right
The response was fast, and speed increasingly determines how these incidents are judged.
PeckShield flagged the anomaly on August 21. The Sandbox disabled bridging to and from Base and BNB Smart Chain, isolating the minted supply and cutting off any route to redeem it through the official bridge.
In its August 22 statement the project said the vulnerability was fully contained, that SAND on Ethereum and Polygon was unaffected, and that no user wallets were compromised.
It advised users not to buy, sell or trade SAND on Base or BNB Chain while liquidity there remains compromised, took a pre-attack snapshot, and said it would compensate eligible liquidity providers based on balances recorded before the attack.
Upbit and Bithumb halted SAND deposits and withdrawals after detecting a possible security incident.
Reading the Project’s Own Statement Correctly
The Sandbox described the impact as less than 0.01% of total SAND supply. That has been read by some as a claim that only a tiny amount of fake SAND was minted.
It is not saying that. It refers to the proportion of legitimate supply affected by the extraction, which is consistent with researchers tracking an enormous quantity of unbacked tokens.
Both statements are accurate about different things. The distinction matters because conflating them makes the project look either dishonest or negligent, and neither reading is supported.
The Context That Sharpens It
SAND’s fully diluted market capitalization sits at roughly $116 million, with the token more than 99% below its 2021 high of $8.40.
Against that base, the drained reserve represents around 0.6% of the entire legitimate token base. A six-figure loss is small next to 2026’s headline exploits and is not small relative to this project.
Optimisus covered the sector decline underneath it in the piece questioning The Sandbox’s future as metaverse hype faded.
SAND rose 4.76% during the episode on trading volume up more than 400%, which says more about speculative reaction to news than about the token’s condition.
Where It Fits in the Year
First-half 2026 security incidents reached 212 exploits, a 13% increase over the second half of 2025, with combined preliminary losses reported above $1.5 billion.
Bridges remain the costliest category within that total.
This one is structurally similar to the Harmony incident Optimisus covered in the piece on 4 billion tokens minted while the supply counter missed it, and to the permission-layer failure in the Maya Protocol six-flaw chain.
The practical takeaway for anyone holding a bridged asset is narrow and useful. A wrapped token is a claim on reserves held elsewhere.
When the minting authority on one side is compromised, that claim stops being backed. The only thing between you and a loss is how fast someone disables the bridge.
Sources
- Crypto.news, The Sandbox’s $49 billion phantom mint, how a bridge exploit created unbacked SAND tokens — https://crypto.news/sandbox-bridge-exploit-49-billion-phantom-sand-mint/
- Crypto.news, SAND bridge exploit contained after unbacked token mint — https://crypto.news/sand-bridge-exploit-contained-unbacked-token-mint/
- The Crypto Times, Sandbox SAND hacked, attackers mint 329 trillion tokens on Base — https://www.cryptotimes.io/2026/08/22/sandbox-sand-hacked-attackers-mint-329-trillion-tokens-on-base-in-5-hour-rampage/
- Blockonomi, The Sandbox bridge exploit, attacker mints 14.9B unbacked tokens on Base network — https://blockonomi.com/the-sandbox-sand-bridge-exploit-attacker-mints-14-9b-unbacked-tokens-on-base-network
- BeInCrypto, The Sandbox contains bridge exploit after unbacked SAND minted on Base and BSC — https://beincrypto.com/sandbox-sand-bridge-exploit-base-bsc/
- Coinpedia, The Sandbox SAND exploit, $49B in new tokens flood Base — https://coinpedia.org/news/the-sandbox-sand-exploit-49b-in-new-tokens-flood-base/
This is not financial advice.
Optimisus covers crypto and technology news for readers who want the detail behind the headline.
