A firmware bug in Coldcard hardware wallets, dating back to March 2021, caused some devices to generate predictable wallet seeds rather than relying on proper hardware randomness, leaving affected private keys guessable. Block Engineering describes the flaw directly, stating that "COLDCARD firmware contains an RNG integration error that causes ngu.random to use MicroPython's deterministic Yasmarang fallback instead of the STM32 hardware RNG."1 Blockhead adds detail on the scope of the weakness, reporting that Coinkite's own technical explanation of the bug says seeds generated on its Mk2 and Mk3 models with the flawed software carried only about 40 bits of real randomness, a small enough set of possibilities that a well-resourced attacker could work through them.
Galaxy Research has tracked the theft in escalating waves. BleepingComputer reports that Galaxy Research "identified an initial wave of transactions that it believes was likely linked to the vulnerability, draining approximately 1,083 BTC, worth $70.2 million, from 1,196 addresses on July 30."2 The losses have grown since then. CoinDesk reports that a fourth wave of sweeps is underway, with researchers estimating the attacker has moved about 1,816 bitcoin, or roughly $114 million, from more than 5,200 addresses since July 30. Fortune's figures align closely, citing Galaxy Research's on-chain analysis showing hackers have moved approximately 1,816 Bitcoin, worth nearly $116 million, off affected wallets. Insurance Journal offers an intermediate snapshot, noting that by Monday, more than 1,755 of the tokens worth some $110 million had been drained from roughly 5,000 wallets, according to Galaxy Research.
Coinkite has released fixed firmware and is urging affected users to migrate their funds to newly generated seeds. Reaction from the security community has been pointed. Decrypt quotes Kraken chief security officer Nick Percoco, who called the incident a "wake-up call for the entire hardware wallet industry."3
For how outlets have framed the story differently, see the framing comparison below.