Blockstream Wrote the Ransom Note 'White-Hat'

The on-chain message left after the drain read: "we are whitehats. contact us on chain."

That is not a security disclosure. That is a hostage note with better branding.

Liquid Network, launched in 2018 by Blockstream, lost roughly 4,000 of its 4,200 bitcoin—about 95% of all BTC pegged to the sidechain. Blockstream attributed it to a software bug in Elements, not compromised keys. The network is now paused.

Translation from Blockstream's incident statement: our code had a bug, someone found it by draining our reserves, and we are asking them politely via blockchain message to give it back.

The Federation Model Has No Recovery Path

This is the part the white-hat framing buries. Liquid uses a federated security model: fifteen functionaries operate the network, while an 11-of-15 quorum controls the Bitcoin peg. Normally, L-BTC must be destroyed before matching BTC can leave federation-controlled reserves. When the code enforcing that threshold has a bug, the attacker does not break through the fence. The attacker becomes the fence.

SideSwap revealed the mechanics: someone exploited a bug in Elements software. "Today at 14:05 UTC, a customer sent 4,000 L-BTC to the SideSwap peg-out service. Our service processed it like any other order: the L-BTC was burned on Liquid with a valid peg-out authorisation, and at 14:28 UTC, the Liquid Federation paid 3,996 BTC to the customer's Bitcoin address. Blockstream has since established that the L-BTC in that order was created through a bug in the Elements software."

Read that again. The federation paid out. The authorization looked valid. Every guard stood down because the code told them to. The federation's 11-of-15 quorum means nothing if the inputs it signs off on can be fabricated at the software layer.

No compromised keys. No stolen hardware. Just a bug that made the accounting lie, and a federation that believed the lie.

The Ledger CTO Is Not a Detail

Ledger CTO Charles Guillemet was skeptical of the white-hat claim. Legitimate security researchers do not drain a bridge and then ask to be contacted on-chain. Withdrawing hundreds of millions of dollars before opening communication differs sharply from conventional vulnerability disclosure practices.

He drew parallels with the Ronin hack (attackers stole around $625 million after compromising validator keys) and the Euler exploit (the attacker sought to negotiate a return). When the CTO of the company that sells hardware wallets to institutional buyers says he does not trust the framing, institutional buyers will not trust it either.

Guillemet suggested these could be "people with good intentions that intensively played with recent LLMs and are not used to responsible disclosures." That is the most charitable read available. It also implies Blockstream's federation reserves are one well-prompted model away from being tested by the next curious researcher.

If you have customer Bitcoin sitting in Liquid right now, you have one operative question: what is the redemption timeline? As of early September, the coins had not left the receiving address, and the sidechain remained paused. The purported white-hat asked whether pushing "most" of the funds back would be acceptable.

Most. That word is doing a lot of work.

The actor is now negotiating to return "most" of the funds to the federation, and the network stays paused. Every hour offline is an hour a competing settlement layer looks more attractive to the 80-plus exchanges that run on Liquid.

What This Is, Plainly

This is not a Bitcoin problem. Bitcoin's base layer was untouched. The lesson applies to any system where code correctness is the sole security boundary: Liquid, other federated bridges, any custody architecture where the smart path through the code is also the only path attackers need.

According to a TRM Labs report, infrastructure and operations vulnerabilities accounted for roughly 15% of crypto incidents during the first half of 2026 but were responsible for 76% of total monetary losses. Low frequency. Maximum damage.

That is the exact risk profile of federated custody, and it just paid out.

The one thing that saves Blockstream's reputation here is a clean return of all funds, a credible post-mortem proving the bug was genuinely non-obvious, and a code audit that closes every related vector before restart. The odds of all three landing cleanly are not high.

But that is Blockstream's problem to solve. Yours is simpler: federated custody gives you the appearance of institutional-grade security right up until the code is wrong. Then the attacker is the institution.

What to watch: Whether the attacker returns all 4,000 BTC or negotiates a de facto bounty; whether federation members publicly debate replacing Elements or auditing its full codebase before restart; whether any exchange announces it is moving settlement volume off Liquid while paused; and how long the pause actually lasts, because every week offline is a free trial for every competitor.