Coldcard Hack Update: $116 Million Swept, but the 15-Attacker Claim Is Still Unconfirmed

Coldcard losses have reached about $116 million, but attribution is still open

About $116 million has been moved from affected Coldcard wallets, but the headline about 15 attackers remains unconfirmed. Investigators have not yet linked this incident to any specific actor. For now, the practical question is not who is behind the exploit, but how much sell pressure may still be left and how much trust damage has already hit Coldcard and Bitcoin self-custody sentiment.

The clearest available picture is that four waves of thefts have touched more than 5,200 addresses, with approximately 1,816 Bitcoin worth nearly $116 million moved from affected wallets. That is the number investors are most likely to price first, because it frames the immediate supply overhang and the scale of the confidence shock.

The speed of the first confirmed sweep also matters. Attackers drained 1,196 Bitcoin addresses in 41 minutes and took 1,082.65 BTC. Galaxy then identified two additional suspected waves of suspicious activity, and the estimated toll kept rising. That pattern suggests an operator trying to liquidate quickly rather than quietly stash funds for later.

The market debate, then, is fairly straightforward. Bulls can argue that much of the damage has already surfaced because so much output has moved. Bears can argue that continued outflows likely point to repeated sells, not a one-time dump. Either way, if more stolen output starts turning into market supply, exposed wallet stockpiles could be discounted before they hit tape.

Block tied the exploit to a 2021 firmware path, not a live device hack

How the vulnerability actually works

This stopped being a generic "offline is safe" story once Block tied the flaw to a March 2021 firmware integration error that routed seed generation to a deterministic software PRNG instead of the hardware RNG. Block said an attacker who can constrain or estimate the device UID, timer state, and prior RNG-call history can reproduce candidate output streams offline, then derive addresses from those candidate seeds and compare them with public blockchain data. In other words, the weakness became an offline guessing exercise against on-chain outputs, not a live intrusion of the device itself.

That is why the issue is not a problem with Bitcoin itself, but a break in the security model at seed creation, where users expect true randomness. If the random path was weak when the wallet was created, the threat remains real even on an air-gapped device.

Coinkite says new seeds are protected, but legacy-seed risk remains

Bears may broadening the argument: if a hardware wallet's randomness stack fails, the "offline is safe" pitch weakens whenever that stack fails. Coinkite's own estimates help explain that concern, with roughly 40 bits of effective entropy on the Mk3 and about 72 bits on the Mk4, Mk5 and Q, versus 128 bits for a 12-word BIP-39 seed.

Bulls have a narrower but valid counter: this is still described as a vendor-specific firmware path, not proof that self-custody fails by default. Coinkite says its patched firmware prevents this issue from affecting any new seed, and it has been explicit that users must generate a new seed and move funds to restore full security. That points to a more targeted repricing: wallet-brand risk and legacy-seed exposure, rather than a collapse of the broader Bitcoin stack.

There is also an important caveat. No public report has yet reconstructed a victim's seed and matched it to a drained address, so the exploit has not been shown end-to-end in public against a confirmed victim wallet.

Even so, investors still have reason to price trust damage. The weakness traces back to a 2021 Coldcard software update, which means affected seeds could have been weakened for years before detection. Coinkite has strongly urged users to move funds as soon as possible, and its own statement made clear how serious the reputational harm is. For investors, the immediate concerns are self-custody sentiment, wallet-brand risk, and the possibility that holders front-run exposure before the market does.

What to do now depends on whether your seed was generated on affected firmware

The positioning call is now practical rather than theoretical. After the full postmortem, investors should have a clearer picture of the true affected inventory, whether the fix is complete, and how much exploit surface remains. Until then, broad portfolio trading still looks premature. The operator has already shown a willingness to keep sweeping after the first losses, and Galaxy's latest read is that the attacker is targeting smaller balances.

Who should move now

Move now if your setup matches the vendor's exposure logic. Seeds created on affected firmware are still at risk unless they were generated with at least 50 independent, private dice rolls or the funded wallet is protected by a strong, unique BIP-39 passphrase. Even then, Coinkite says a weak, uncertain, or reused passphrase does not truly remove exposure, and a strong passphrase does not repair an affected seed. Unless the dice-roll exception clearly applies, the recommended action is to replace the seed and migrate as soon as practical.

Who should wait

Wait if you are unsure how your seed was created, cannot verify the dice-roll exception, or are considering a broader portfolio response beyond your own Coldcard exposure. The final call still depends on the full postmortem. Investors still need the confirmed affected list, proof that the patch stops the issue, and clarity on remaining edge cases before rewarding or punishing broader self-custody exposure.

Firmware threshold list

Before generating any new seed, confirm the device is patched to at least:

• Mk2/Mk3: firmware 4.2.0 or later
• Mk4/Mk5 standard: firmware 5.6.0 or later
• Q standard: firmware 1.5.0Q or later
• Mk4/Mk5 Edge: firmware 6.6.0X or later
• Q Edge: firmware 6.6.0QX or later

Standard and Edge are separate release tracks, so an older Edge build should not be treated as safe just because its version number looks higher than a standard release.

What to watch next

If you are unsure how your seed was generated, proceed carefully. A mistake during backup, passphrase entry, or the test transaction can be just as costly as the exploit itself. The next useful updates are the full postmortem, the confirmed scope of affected inventory, and clear evidence that new seeds generated on patched firmware are no longer vulnerable.