What happened
On 15 November 2026, Luxor Technology said its monitoring detected unauthorised changes made through Commander, its remote mining-fleet management service. The changes pointed machines belonging to just over a dozen client accounts at another pool and redirected the resulting rewards.
Luxor said it contained the path and patched Commander within two hours. The company also said it began restoring affected machines, found no data exfiltration, and would compensate affected clients in full. Luxor Pool and the LuxOS mining firmware were not affected, according to the company.
The Block obtained a direct estimate from Luxor chief executive Nick Hansen that about 13 exahashes per second were affected, representing roughly 20,000 US dollars of revenue per hour at the time. That is a company estimate of hashrate exposure, not a final accounting of losses.
What it changes
Commander is a control plane. It lets an operator configure many miners without visiting each machine. That saves labour, but it also means an unauthorised change can redirect a large fleet at once. The machines can keep hashing and consuming electricity while someone else receives the pool credit.
CoinDesk independently checked the reported Bitcoin address and counted three OCEAN block-reward payments totalling about 0.243 BTC by Monday morning. Cointelegraph had observed two payments totalling about 0.162 BTC at an earlier snapshot. The different totals reflect observation times, not a contradiction about the same fixed balance.
Those payments can arrive after the miners stop pointing at OCEAN. Under OCEAN's TIDES payout system, submitted shares remain in a rolling eligibility window. A miner can therefore receive a later block payout for work submitted before the redirect was stopped. The address balance is evidence of pool rewards, but it is not by itself a complete measure of the incident.
For operators, the practical response is to treat remote fleet management as privileged infrastructure. Access should be limited, configuration changes should generate independent alerts, and pool destinations and payout addresses should be monitored outside the same management system. Luxor has not yet published enough technical detail to identify a specific customer-side control that would have prevented this incident.
What it does not change
The incident does not show a break in Bitcoin's consensus rules, proof of work or mining hardware. The miners continued producing valid work. The failure was in who controlled their pool destination and who received credit for that work.
Luxor says the compromise did not reach Luxor Pool or LuxOS, and that client data was not exfiltrated. Those statements have not yet been supported by a public technical postmortem. They define the company's current scope assessment, not an independent audit.
Hansen described the event as an AI-enabled attack. None of the public reports explains what an AI system did, how access was first obtained or which Commander component failed. Decrypt and Bitcoin.com News both note that the root cause and claimed AI role remain undisclosed. Until the promised postmortem appears, that label should not be treated as a demonstrated mechanism.
Early public alerts also estimated that 64,000 machines were involved. Luxor confirmed just over a dozen affected accounts and gave The Block the 13 EH/s estimate, but it has not publicly confirmed the machine count. Account count, machine count and hashrate are different measures.
Context
Three independent newsrooms added distinct reporting. The Block obtained the company's hashrate and revenue estimates. CoinDesk assessed the reported address and OCEAN payout records. Cointelegraph contacted Luxor and checked the address at an earlier point. Decrypt and Bitcoin.com News provided additional review of the company statement and its unanswered technical questions.
The reports agree on the affected service, approximate account scope, two-hour patch window, compensation promise and separation from Luxor Pool and LuxOS. They also agree that no public root-cause analysis is available. The main uncertainty is not whether rewards were redirected, but how access was obtained, what controls failed and whether Luxor's current scope changes after forensic review.
A promised postmortem could answer those questions. Until then, the defensible conclusion is narrower: a central fleet-management service became a high-leverage point for redirecting otherwise valid Bitcoin mining work.
