BTCPay Server Orders Emergency Shutdown After Hot Wallets Swept
Merchants urged to take payment nodes offline immediately as founder Nicolas Dorier links compromises to pre-2.4.2 deployments; exploit underscores the operational boundary of self-custody.
An emergency advisory issued Sunday evening sent shockwaves through the self-hosted payment ecosystem as merchants and node operators were urged to shut down active deployments without delay. We are receiving reports of hot wallets on BTCPay Servers being swept by unauthorized attackers. If you're running an instance, please shut it down (./btcpay-down.sh) until we complete the investigation.
warned the project's core team [s580]. Core contributor Uncle Rockstar Developer underscored the urgency: Not a drill: shut down your BTCPay Server if you have a hot wallet.
[s580]
Merchants moved quickly to bring infrastructure dark. SweetSats confirmed its payment gateway was suspended: PSA: Our BTCPayServer on SweetSats.io is currently shut down until further notice. 🐝
[s61] Operator Nodestrich echoed the alarm: Very serious incident: If you are a BTCPay Server user, please be advised there is an attack under way. ⚠️
[s116] As Psyduck also sounded the alert across social feeds [s19], technical details began to surface from project maintainers.
Preliminary findings shared by founder Nicolas Dorier pointed toward instances running releases prior to version 2.4.2, where attackers obtained administrator access on affected hosts [s580]. Version 2.4.2 had been issued in August to address a vulnerability targeting Lightning nodes, yet the project stopped short of confirming newer builds were unaffected, maintaining a blanket shutdown recommendation. The breach brought renewed scrutiny to the mechanics of automated hot wallets, where signing seeds reside directly on network-connected servers [s580]. As analysts noted in retrospective reviews of the exploit, watch-only configurations where keys reside on offline hardware signers remained isolated from exposure [s580]. Self-custody closes the counterparty risk. It does not close the operational risk. Those are different problems.
observed an incident review circulating across long-form feeds [s580].