Coldcard's 72-Bit Flaw Just Cost Bitcoin $89M-Why the Real Risk Is Still Unseen

The exploit is about weak seeds, not user error

A severe entropy collapse in Coldcard firmware has been linked to roughly $89 million in Bitcoin losses, and the attack may still be active. By early August, researchers had tied roughly 1,367 BTC across 4,585 addresses to the flaw. Coinkite disclosed the issue on July 30, and the first visible sweep hit about 500 single-signature addresses the next morning. The funds were taken without phishing, malware, or physical access to the devices.

The negative implication is straightforward: even premium Bitcoin custody hardware can carry a randomness weakness for years, and updating the firmware does not fix a seed already created weakly. The more constructive take is narrower but real: the open-source nature of Bitcoin wallet code helped the community detect and respond quickly, and Block, Trezor, and Ledger have said their products are unaffected.

For now, the key question is containment. If fresh waves of sweeps keep showing up, investors are more likely to treat Coldcard exposure as an active loss pool rather than a closed bug report. If the pattern stops spreading, the episode can more quickly be reclassified as an isolated wallet incident.

How Coldcard's firmware turned seed generation into a guessable-key problem

This was a firmware integration error inside the wallet, not a blockchain settlement issue or a network-level exploit. In March 2021, Coldcard firmware was changed so that seed generation was routed to a deterministic software pseudorandom number generator instead of the STM32 hardware random number generator. That matters because a seed is only as strong as its randomness.

The entropy drop matters more than the headline says

A normal 12-word BIP-39 seed is meant to carry 128 bits of entropy. On Coldcard Mk3 devices, Coinkite estimated effective entropy fell to roughly 40 bits. On Mk4, Mk5, and Q models, it was still severely weakened at about 72 bits. Those figures move the problem out of the realm of infeasible brute force and into the realm of a targeted offline search.

Why attackers did not need to touch the device

Block showed that if an attacker can determine or sufficiently constrain the device UID, timer state, and prior RNG-call history, they can reproduce candidate output streams offline without accessing the device at all. Once those streams are narrowed, the attack is straightforward: derive addresses from candidate seeds and compare them with public blockchain data. The wallet did not need to be hacked in the traditional sense; the weak seed only needed to be predictable enough for an offline guess to match something already visible on-chain.

Coinkite shipped a fix on July 31, but that patch only prevents future weak generation on patched firmware. It does not repair an existing seed. Restoring or copying a weak seed preserves the exposure: if an old seed came from an affected Coldcard before the fix, putting it back into updated firmware or another wallet reuses the same guessable key material.

What owners should do, and what would calm the market

The only instruction that matters right now is Coinkite's: owners of affected devices should generate a new seed on patched firmware and move their coins. This is not a broad Bitcoin protocol alert. It is a wallet-specific key-management event.

Who still needs to act

If you used an affected Coldcard device before the fix, the priority is simple: - Create a fresh seed on patched firmware. - Move funds from any wallet derived from the old, potentially weak seed. - Do not restore the old seed into updated firmware or another wallet, because that carries the weakness forward.

Why this has not become a broader Bitcoin panic

This matters for exposed wallets, not for the wider custody stack. Block, Trezor, and Ledger have confirmed their products are unaffected, and no public report has reconstructed a victim's seed and matched it to a drained address. That does not make the incident less serious, but it does mean the damage appears tied to known weak-seed behavior rather than to the exposure of individual victim secrets.

What would reduce the risk narrative

The clearest reassurance signal is simple: no new waves, no new clusters of affected addresses, and no spillover to other wallets. Coinkite's fix shipped on July 31, and roughly 1,367 BTC across 4,585 addresses had already been linked by August 1, so the situation still deserves caution rather than relief. The real warning sign is not negative press for Coldcard; it is fresh sweeps, broader wallet exposure, or any attempt to link reconstructed seeds directly back to specific victims.