Coldcard's $38M BTC Sweep: Who Lost Funds, Who's Still Exposed, and What to Watch Now

What happened in the Coldcard exploit

A rapid sweep from weak wallet seeds

Earlier today, 594 BTC worth about $38 million was swept from around 500 separate Bitcoin wallets in a 25-minute window. The speed and pattern of the drains suggest the attacker was targeting wallets tied to specific creation dates, rather than randomly scanning the chain.

This was not a Bitcoin protocol failure. It was a device-level seed-generation failure. A faulty random-number generator meant some devices did not use the intended randomness source during wallet creation, making some seeds easier to guess than they should have been. That strikes at the core job of a hardware wallet: keeping private keys unguessable.

Coinkite has released fixed firmware, but it is also clear that updating the firmware does not repair a seed. Affected users need to generate a fresh seed on an updated device and move funds there.

Who is exposed and what to do next

The highest-risk group is narrower than the panic implies. If a seed was created on an Mk3 running firmware 4.0.1 through 4.1.9 without at least 50 independent, private dice rolls, treat that wallet as exposed and migrate it promptly. That matches the profile behind the earlier sweep, where around 500 separate Bitcoin wallets were drained in a 25-minute window.

Who should move funds now

Do not wait to see whether there is a second sweep. If you are in that Mk3 cohort, the cleaner path is:

1. Update the Mk3 to firmware version 4.2.0 or later before creating a replacement wallet.
2. Generate a completely new seed on the updated device, then record and verify the new backup, wallet fingerprint, and a receive address.
3. Send a small test transaction first, then move the remaining funds.
4. Keep the old backup until the migration is fully confirmed.

That sequence matters because updating firmware does not fix an already-generated weak seed. Only a new seed and a fund move close the gap.

Where Coinkite draws the line

For seeds created on later Coldcard firmware, Coinkite's guidance is more conditional. If you added at least 50 fair, independent, private dice rolls when the original seed was generated, Coinkite does not consider that seed at risk from this RNG issue alone. Still, review the advisory's dice guidance before migrating, because the workflow can differ.

This is not a verdict on every self-custody setup. It is a firmware-specific issue tied to seed creation on affected devices and firmware ranges. For newer devices, Coinkite says users on affected Mk4, Mk5, or Q firmware should still upgrade to the fixed firmware for their track before generating a new seed.

Why the bigger question is custody trust, not Bitcoin itself

The market read-through is less about Bitcoin's fundamentals and more about confidence in this incident's hardware-wallet maker. The key tell is that every drained wallet was single-signature. That keeps the debate focused on which custody workflows users will trust after a device-level failure.

Where the pressure lands first

The first pressure point is brand trust. Users now have a concrete reason to ask whether a single-device workflow is resilient enough, especially while migration steps, replacement demand, and support demand converge. That matters because top cold wallets compete on security and trust first, features second.

What to watch next

Watch whether this remains a Coldcard-specific event or becomes a broader shift in how users think about single-key hardware wallets.

• Bull view: the damage stays contained to Coldcard and fades once migration stabilizes.
• Bear view: the incident becomes a larger signal that users should prefer multi-key or alternative custody designs.

The next few weeks on migration progress, supply, and user substitution will do more than any headline to show which trajectory is happening.