NEAR Intents confirmed an exploit of more than $3.8 million on October 1, 2026, closed the vulnerability and halted deposits and withdrawals on eleven networks. Full reimbursement of all affe
NEAR Intents confirmed an exploit of more than $3.8 million on October 1, 2026, closed the vulnerability and halted deposits and withdrawals on eleven networks. Full reimbursement of all affected funds has been promised; a date for it has not. If you have used the swap layer to move balances between two blockchains, one thing now decides above all: where your balance sits at this moment, and whether it has to stay there.
The price of NEAR stood at $4.90 on the afternoon of October 1, 2026, 6.96 percent below the previous day; over the week the value is up 7.66 percent. The price here is only the visible part, though. The more important part is an infrastructure through which, according to the operators, more than $30 billion has already flowed across 35 networks, and which was only partly usable for several hours.
NEAR Intents is the cross-chain swap layer in the orbit of the NEAR protocol. This layer takes in an intention, such as swapping a token on one blockchain for a token on another, and lets service providers compete to execute it. On Thursday, irregular withdrawals flowed out through a hot wallet of this layer. The damage, as the project describes it, amounts to more than $3.8 million.
The team publicly acknowledged the incident, closed the affected point in the smart contract and suspended deposits and withdrawals on several connected networks. Core operations were, by its own account, to resume quickly, while deposits and withdrawals stay down longer. That order matters more to you than the damage figure: a service that permits swapping but not withdrawing is not usable for you.
A hot wallet is a wallet whose keys sit on a system connected to the internet, so that a platform can execute payments automatically. That access is the reason it is fast, and at the same time the reason it remains the preferred target of an attack. Set against it is the cold wallet, whose keys stay offline.
The project describes the cause as a fault in the way the Omni system for deposits and withdrawals interacted with the NEAR Intents smart contract. It was not a single component that was defective, but the handover between two. That is the normal case in attacks on cross-chain infrastructure, and the reason audit reports on individual contracts say only so much: what is audited is usually the building block, what is exploited is the joint.
In this architecture Omni is the layer that accepts deposits from an external blockchain and releases withdrawals back to it. The Intents level above it decides what happens to the balance. Anyone who can get the order of the two levels out of step can trigger a withdrawal for which there was no valid deposit. The contract-side gap has been closed, according to the team.
A closed contract fault prevents a repeat of the same attack. It says nothing about whether the same joint is still open elsewhere, and it does not bring back any funds that have flowed out. What counts for you is therefore not the notice about the patch but the release of withdrawals on the network where your balance sits.
Affected by the suspension, according to the CoinDesk report, were BNB Smart Chain, Polygon, TON, Optimism, Avalanche, Stellar, Monad, X Layer, ADI, Scroll and Plasma. This list is the day's actual finding. A single faulty handover point can shut down access to eleven different ecosystems at once, and the list shows at the same time how many chains now hang off a shared layer.
In practice that means a swap in progress whose counter-value was due to arrive on one of these networks could be left hanging. A balance already sitting on one of these chains was not automatically affected, as long as it was in a wallet of your own and not in the swap layer. The distinction between in my wallet and in transit within a service decides the damage in hours like these.
Eleven networks hang off the same deposit and withdrawal layer: if it tears, all of them stand still.
The investigator ZachXBT, who publicly analyses on-chain traces, has retraced the movement: the irregular outflows began at a hot wallet on BNB Chain attributed to NEAR Intents. From there the funds went to the KuCoin exchange and were swapped into bitcoin. This chain is typical, because bitcoin offers the deepest liquidity and a move through an exchange breaks the trail as soon as accounts are interposed there.
The project says it has reported the incident to law enforcement and brought in security and analytics firms to follow the funds further. Whether an exchange freezes the amounts received is its own decision, and usually only after a formal request. There are no reliable statements about recovery on this day, and nobody should promise you any.
The team has promised to reimburse affected funds in full. No timetable was named. This combination is the most delicate point of the incident, because it creates a claim you can neither quantify nor date: your claim is against a project, not against a supervised institution with deposit protection.
In practice that means three things. Your balance in the swap layer is unavailable until release. A promised reimbursement without a date is a declaration of intent, not a due date. And as long as you have no evidence of the state of your balance before the incident, you carry the burden of proof for your own claim. A screenshot with a date and the transaction ID of the last swap are worth more at this point than any market analysis.
Anyone holding long-term positions should take the opportunity and put them where no third-party service stands between them and the key. Which devices and programmes come into question for that and how they differ is set out in the hardware wallet comparison. A swap layer is a passage, not a warehouse.
An intent is a declaration of intention: you set out what you want to give up and what you want to receive, and leave the route there to others. A solver is the service provider that carries out this intention and earns on the price difference. A classic bridge, by contrast, locks your token on one chain and issues a representation of it on the other.
The difference is decisive for the risk question. With a bridge, the risk lies in the locked holding that backs the representations. With an intent system it lies in the deposit and withdrawal layer and in the hands of the solvers, who temporarily have other people's balances at their disposal. That is precisely the layer affected on October 1. For you that means the question is not whether a method is safe, but how long your balance is in a third party's hands at all.
A swap completed in seconds exposes you briefly to a failure. A swap whose counter-value arrives only minutes or hours later exposes you for a long time. A balance you leave sitting in the service after the swap exposes you permanently. The third variant is the most expensive and at the same time the most common, because it is convenient.
The same platform was on the other side of events the week before: back then the service intercepted some $50 million from the Bitget hack before the money could move on. We described that on September 30, 2026, under the title "NEAR Intents blocks $50 million from the Bitget hack". Seven days later, $3.8 million is missing from the same infrastructure.
No schadenfreude follows from that, but a sober insight. A platform able to stop other people's funds has deep insight into payment flows and correspondingly many points of contact. The same reach that makes a block possible creates the attack surface. Anyone who looks only at a service's capabilities and not at the number of its handover points is pricing the risk too cheaply.
Reimbursement is promised, a date for it is not: for those affected, time is running without a date.
Four concrete points are at stake, and none of them takes more than a few minutes.
What matters is the difference between waiting and doing nothing. Waiting means watching the situation, securing evidence, putting no new funds into the affected layer. Doing nothing means leaving a balance there and being unable to show later how much it was.
Since January 1, 2026, only authorised providers may offer crypto-asset services in Germany; the national transition periods in the EU expired on July 1, 2026 at the latest. A decentralised swap layer is not a supervised custodian, and that is exactly what the legal consequence hangs on: there is no deposit protection, no supervisory complaint to BaFin over a duty to provide a service, and no body that enforces a reimbursement.
Anyone wanting to draw this line cleanly in daily use buys and sells through an authorised provider and uses cross-chain routes only for the purpose they were built for. An overview of providers with European authorisation is in the hub on regulated crypto exchanges. That replaces no judgement of your own, but it moves the part of your assets entrusted to a third party into a framework with obligations.
The one-year holding period under German income tax law attaches to the acquisition and disposal of the same asset. Whether a reimbursement counts for tax as a reversal or as a new acquisition depends on how it is technically executed: whether the same token comes back or a counter-value in another currency. You should document that distinction before the reimbursement happens, not after.
The NEAR price stood at $4.90 on the afternoon of October 1, 2026, after a day's high of $5.52 and a day's low of $4.79. The loss of 6.96 percent against the previous day makes NEAR the weakest value among the 25 largest cryptocurrencies that day; the market capitalisation is around $6.4 billion, daily turnover some $1.4 billion. Over seven days a gain of 7.66 percent still stands. The figures come from CoinGecko.
These numbers describe a reaction, not a valuation. A decline of almost 7 percent on damage of $3.8 million shows that the market classes the sum itself as small and the interruption of the swap layer as the costlier part. How things go from here depends on how quickly deposits and withdrawals run again on all eleven networks.
The incident does not stand alone. On September 30, 2026 we compiled the finding that crypto hacks caused losses of $766 million within a month. Against that backdrop $3.8 million is a small item, and for that very reason the lesson from it is the more important one: it is not only the large sums that are hit, and not only the unknown projects.
For a German tax return, what you can evidence is what counts. Secure the following now rather than later: the holding before the incident with its date, the transaction IDs of the affected operations, the project's public statement with its date, and every later credit with its amount and time. A loss that cannot be evidenced has no effect for tax, and a reimbursement whose origin you cannot explain prompts questions.
Whether a loss from a theft is deductible at all has not been conclusively settled in Germany and depends on the individual case; only a tax adviser or the tax office can give a binding answer on that. The first step is undisputed: a complete, timely record. How the basic rules on holding period, allowance and reporting obligations interact is set out in our overview of crypto tax in Germany.
Sources for further reading: the report by CoinDesk of October 1, 2026 with the list of affected networks and ZachXBT's findings, and the NEAR Intents documentation on how intents and solvers work.
(As of October 1, 2026. This article is not investment advice. Prices and fee structures change; check the terms with the provider before you buy.)
