Coldcard Flaw Costs Bitcoin Users $89 Million - Is the Next Target Your Wallet?

The Coldcard exploit is still active, not just a past bug

This is no longer only a historical security incident. Recent onchain sweeps suggest the extraction is still happening, and the total exposure keeps growing.

July 30 showed how fast the attacker could move

On July 30, attackers swept 1,196 Bitcoin addresses in 41 minutes, taking 1,082.65 BTC worth about $70.2 million. That pace suggests automated harvesting across a broad set of weak addresses rather than slow, targeted picking of individual wallets. For anyone using an affected Coldcard setup, delay increases the chance that exposure is already actionable.

Losses climbed as Galaxy identified two more suspected waves

Galaxy later tied two additional waves to the same weak-key problem, lifting the toll to nearly $89 million across three suspected waves. Block said the emergency firmware shipped on July 31, but the advisory was clear: installing the update does not repair a seed that was already generated on the vulnerable path. The patch prevents further weak-seed creation; it does not undo damage already done.

Are the attackers contained, or still searching?

That is the live question. One side of the debate argues the exploit is mechanical and therefore containable: once weak seeds are found and drained, the opportunity shrinks. The other side argues the threat is still active because Galaxy flagged a third wave of sweeps. For users, the practical issue is not whether the bug is historically resolved. It is whether the attacker is still finding and harvesting weak seeds faster than affected holders can respond.

Why a 2021 Coldcard firmware mistake still matters today

The clearest way to understand this incident is as a key-space compression problem, not a traditional remote hack.

What actually broke

A March 2021 firmware integration error routed seed generation to a deterministic software pseudorandom number generator instead of the STM32 hardware random number generator. In practical terms, the weak link is no longer "someone breaking into your device." The risk is an attacker reproducing candidate seed streams offline, deriving addresses from them, and checking those addresses against public blockchain data.

Why the entropy gap matters

According to Coinkite, effective entropy dropped to roughly 40 bits on the Mk3 and about 72 bits on the Mk4, Mk5 and Q, versus the 128-bit security investors expect from a 12-word BIP-39 seed. That gap turns brute-force search into a feasible offline candidate-filtering problem for anyone who can constrain the device UID, timer state, and prior RNG history. The important detail for users is simple: exposure depends on the firmware state when the seed was created, not just what is installed today.

Why attribution is still cautious

Skeptics have a fair point: no public report has reconstructed a victim's seed and matched it to a drained address. That means the case is better described as a demonstrated vulnerability linked to theft, rather than a fully proven end-to-end attribution chain. Even so, that distinction should not dictate user behavior. If your wallet was generated under the bad path, the cost of being wrong is permanent loss, while the cost of migrating is mainly time and fees.

The current watchpoint: wave three changed tactics

Wave three also changed the signal profile. Galaxy flagged a third wave of sweeps tied to weak Coldcard-generated keys, with activity targeting smaller balances and changing how funds are collected onchain. If collections are becoming smaller and more fragmented, that may signal a broader sweep rather than a closing window of opportunity.

> If you used an affected Coldcard model and firmware before the fix, treat this as actionable now and migrate funds under patched firmware.

What to do if you may be affected

This looks more like a wallet-specific issue than a Bitcoin macro shock

Losses across the three suspected waves total 1,367 bitcoin from 4,585 addresses. Relative to the total Bitcoin supply, that is small. That is not, by itself, a case for broad BTC liquidation. But it is large enough to matter immediately for exposed Coldcard holders.

The cleaner response is migration, not panic selling

Recent activity already showed the threat is live, from the initial more than 1,000 bitcoin from 1,196 digital wallets in just 41 minutes to a third wave of sweeps tied to weak Coldcard-generated keys. If you fall in the affected set, the more measured move is to move funds under patched firmware and avoid restoring the old seed into updated software.

What would weaken the urgency?

This situation becomes less urgent if it remains a narrow Coldcard trust issue rather than spreading further. The case for immediate wallet-level risk reduction weakens if new sweeps stop, advisories do not broaden, and the weak-key pattern stays confined to the already identified firmware flaw in Coldcard. If, instead, new sweeps keep appearing or the pattern spreads beyond Coldcard, the risk profile would need to be reassessed.