Coldcard Losses May Hit $130M as 15 Attackers Hit at Once

Galaxy's $130 million estimate shifts the incident from scare to active risk

Galaxy's latest estimate turns this from a security alert into a live liquidity event. The firm has confirmed 1,596 BTC stolen from ~7,300 addresses across three attack waves, and it flags a suspected fourth wave that could bring the total to 2,055 BTC, or roughly $130 million.

That matters because the exploit did not require one patient actor. At least 15 different attackers have been linked to the incident after new victim reports surfaced. That breadth suggests the vulnerability remained exploitable by multiple parties, not just one coordinated group.

Most Bitcoin is still safely held, but investors cannot assume the exposure is fully contained. Galaxy has cautioned that not every affected wallet was created using the vulnerable software, which means the incident is tied to specific wallet-generation conditions rather than simply to ownership of the hardware. As long as vulnerable seeds remain in use, the risk of additional losses stays more serious than a one-hour headline.

A 2021 firmware change created a reusable offline attack path

How the vulnerability worked

The root cause goes back to a single code change made on March 1, 2021 in Coldcard firmware. That change caused seed generation to fall back to a software-based pseudorandom number generator instead of the device's hardware random number generator. The result was far less entropy than the design intended.

That reduction is what kept the attack relevant years later. Block estimated that the effective search space collapsed to roughly 40 bits under some conditions on Mk3 devices, while Mk4, Mk5, and Coldcard Q devices remained below 73.3 bits instead of the intended 128-bit security level. In practice, that means an attacker does not need physical access to the device; they can generate candidate seeds and test them against public blockchain data.

Why multiple attackers could exploit the same flaw

This is why the attacker count matters. Galaxy said 15 different attackers have been linked to the exploit, which fits a weakness that can be monetized with offline computation rather than one that requires a single sophisticated operation.

The suspected fourth wave also looked different from a closed post-mortem. Galaxy observed 218 transactions across 462 victim addresses in blocks 960,778 through 960,792, nearly 45 times the prior background rate. The same report said the activity reportedly used Bitcoin's replace-by-fee mechanism, leaving some transactions unconfirmed and potentially contestable before settlement.

What still is not fully confirmed

Some parts of the exposure map remain provisional. Galaxy has kept the suspected fourth wave out of its confirmed figures because no victims have yet reported being part of it. Researchers have also warned that not every affected wallet was created using the vulnerable software, so ownership of a Coldcard alone does not automatically mean a wallet was generated with the weakened randomness.

That leaves a clear split in the read-through: users can reduce future risk by identifying affected devices and applying updates, but wallets created with vulnerable seeds remain exposed regardless of current device firmware.

The next market test is trust in self-custody, not just the final loss count

The next repricing is less likely to come from another update to the loss tally and more likely to come from how investors view self-custody risk after this event. This incident showed wallet-layer compromise can create losses without a centralized exchange failure, which cuts to the value proposition of hardware wallets.

Signals to watch

• Whether affected users move funds away from store bitcoin offline devices and back toward hosted custody.
• Whether exchange deposits rise from recently affected addresses rather than from routine wallet activity alone.
• Whether adoption of the fix expands beyond the initial affected cohort.

What would ease pressure

A quieter attack pattern would help the market calm down. Galaxy has relied on new victim reports to uncover additional attacks, so fewer fresh reports would suggest the exploit is becoming harder to surface. Confirmation that affected users are moving funds away from the vulnerable seed path would also improve the read.

What would keep the stress alive

If reports continue to show new exploitation, custody risk remains live. More broadly, if investors begin reevaluating hardware wallets after a years-old randomness flaw turned into a large-scale loss event, the damage will extend well beyond the wallets directly affected. For now, the key question is not only how much BTC was taken, but whether confidence in self-custody tools has been meaningfully weakened.