# Solflare freezes staking rewards after external provider failure

> An incident at a staking provider leaves Solflare's validator without rewards. User funds are safe and there will be compensation.
> By José Maldonado · DeFi · 2026-10-29

Users who delegate to Solflare's validator have gone two days without receiving rewards, and the reason is not on-chain but at a provider. The non-custodial Solana wallet announced on October 27 that a security incident at one of its staking providers affected private keys and left its validator unable to generate or distribute rewards. The company has stressed that user funds are not part of the damage.

The detail that makes this case more than a service notice is that: the problem was not at Solflare, nor in a contract, nor in the network. It was at a third party that held keys. And when a third party holds keys, the user of a non-custodial wallet discovers that their custody was not, in fact, custody. Delegation, in practice, is a trust chain with several layers and only one of them appears in the interface.

Solflare has [said](https://x.com/solflare/status/2115223521395896530) it plans to compensate those who migrate to a new validator. It is the expected response and it leaves the same two questions open: how long does the compensation take to arrive, and what guarantees are there that the same scheme will not be repeated?

## What does delegating to a validator mean?

In a proof-of-stake network, the user who delegates does not hand over their coins: they assign them to a validator so that it produces blocks on their behalf. The delegator retains ownership and receives a share of the rewards that validator generates.

The model has an uncomfortable requirement. To operate, the validator needs keys and infrastructure that is often outsourced. That is where staking providers come in, companies that manage nodes, sign transactions and keep secrets on the operator's behalf.

When the provider has an incident, the validator stops producing and rewards halt. The coins stay where they were, and the income stream that justified delegating disappears until it is resolved.

## The link that is not visible

For the user, the path of a delegation seems short: open the wallet, choose a validator, confirm. What lies behind it is a trust chain with several layers, and almost none of them is shown in the interface.

The first layer is the wallet, which in Solflare's case does not custody the funds. The second is the validator, which may belong to the company itself or to a partner.

![Solflare reports the event on its X account](https://content.observatorioblockchain.com/wp-content/uploads/2026/10/solflare-wallet-x.webp)
*Solflare reports the event on its X account*

The third is the infrastructure provider, which is where this has happened. The fourth is the keys that provider handles, and that is where a technical incident becomes a trust problem.

That scheme is not exclusive to Solana or to Solflare. It is the standard model of institutional and professional staking, and it works well as long as no one loses a key. This week's incident is a reminder that the decentralization of a network is not automatically inherited in the operation of its participants.

Layer

Who is responsible

What has failed here

Wallet

Solflare

Nothing: it does not custody the funds

Validator

Solflare

It stops generating and distributing rewards

Staking provider

Unnamed third party

Security incident with private keys

Delegated funds

User

Not affected according to the company

## Questions left unanswered for the community

The first is exactly which keys were affected. Solflare has spoken of private keys held by the provider, and the difference between a block-signing key and a withdrawal key is enormous in terms of risk. Without that detail, the user cannot know whether their delegation is exposed or merely interrupted.

The second is the name of the provider. The company has not made it public, which is common during an investigation and also prevents other operators using the same service from acting in advance.

The third is compensation. "Compensate those who migrate to a new validator" is a phrase that needs numbers: what it is calculated on, from what date and within what timeframe. As long as there is no figure, it is an intention.

## Staking still has these scares

Staking is sold as a form of passive yield and is, in reality, an operation with counterparties. The user does not lend money to a fund, but does trust that a series of technical actors keep their systems healthy and their keys safe.

That trust is paid for in basis points of yield. The validator offering the lowest fee is usually the one that outsources the most, because in-house infrastructure costs money. The user who optimizes solely for profitability is buying the most fragile link in the chain.

The reasonable alternative is not to stop delegating, but to look at who is behind it. A validator with its own infrastructure, a public track record and a somewhat higher fee is usually a better deal than one that outsources everything to cut costs. The difference in annual yield rarely makes up for a month without rewards.

## The parallel with other gaps

The two incidents this week point to the same pattern. The theft of 9,408 SOL across forty-eight wallets from the [Bloom trading bot](https://observatorioblockchain.com/defi/bloom-bot-9408-sol-94-segundos-solana/) also occurred in the layer that signs on the user's behalf, not on the chain. And the case of the [credentials stolen by Aokigahara](https://observatorioblockchain.com/ciberseguridad/aokigahara-infostealer-tokens-ia-desarrolladores/) shows the same thing from the desktop: the weak link is in the software that has permission to act in your name.

The common lesson is that the security of an asset depends on every entity that can move it or sign for it. In a decentralized network, that list of entities can be longer than the interface suggests.

## Compensation as proof

Promising compensation is easy; delivering it is what separates a serious operator from one that improvises. The company has tied the payment to the migration to a new validator, a formula that solves two things at once: it compensates the user and rebuilds the delegation on different infrastructure.

The missing detail is the calculation method. If compensation is based on forgone rewards, a reference rate and a time window must be set; if it is a fixed amount, the user with more delegated ends up worse off. Neither option is wrong, and both must be written down before they are announced.

## What can a delegator do?

The first thing is to review who you delegate to and how much of its own infrastructure that validator has. Second is to diversify: spreading the delegation across several validators reduces the impact of an individual incident, though it does not eliminate it.

The third is to distrust profitability as the only criterion. A validator promising more than average is taking on a risk that someone pays for, and that someone is usually the delegator when something fails. It is worth reviewing every quarter where the delegation sits, because operators change providers without telling their users.

Solflare did the right thing by communicating quickly and separating the damage from the funds. What remains to be seen is whether the incident is closed with compensation that has a date and with a technical explanation from the provider. Until then, the validator will remain halted and the conversation about who has custody of what in staking will remain an unresolved issue.
