Liquid recovers 3,400 BTC; 598.5 BTC still outside

Liquid’s federation wallet received 3,400 BTC back on September 7, one day after self-described white-hat hackers withdrew roughly 4,000 BTC, but holders, exchanges and market makers remain unable to rely on normal settlement while about 598.5 BTC is outstanding and the network is paused. The transfer returned about 85% of the bitcoin removed. It reduced the immediate collateral deficit, but it did not repair the process failure that let newly created, unbacked L-BTC pass through a standard peg-out route and become real bitcoin on the base chain.

How did the Liquid Network exploit work?

The withdrawal succeeded because a consensus bug in Elements, the software beneath Liquid, allowed the actors to create approximately 4,000 L-BTC without a matching bitcoin deposit. They sent those coins to SideSwap’s peg-out service on September 6. SideSwap burned the L-BTC with valid authorization, and the federation’s signers released 3,996.02 BTC; the service forwarded almost 3,996 BTC to the supplied address in the same Bitcoin block.

• The unbacked L-BTC was created at 13:53 UTC.
• A 4,000 L-BTC peg-out order arrived at 14:05 UTC.
• The federation payout followed at 14:28 UTC, a 23-minute path from order to bitcoin release.
• SideSwap charged its usual 0.1% fee, about 4 BTC, and later returned it to the federation.

No SideSwap key was reported compromised. The operational failure was broader: an online authorization key, automatic forwarding and no size, velocity or wallet-age controls let an order representing most of the reserve clear without human review. Eleven of 15 federation signatures were also supplied without stopping it.

How much bitcoin is still missing from Liquid?

About 598.5 BTC remains outside the federation wallet, while Blockstream continues discussions with the actors. Before the incident, the wallet held roughly 4,200 BTC; the withdrawal left about 197 BTC, so the 3,400 BTC return restores most, not all, of the reserve. Those figures are rounded public balances rather than a full liability audit, and they do not establish that every legitimate L-BTC claim is presently redeemable.

The observable record shows the return followed Blockstream’s notice that affected bridge nodes had been patched. It does not prove the actors’ motives or make the “white hat” label settled fact. Nor has the remaining amount been publicly established as an agreed bounty. Until it returns or the federation covers the gap, users bear counterparty-style backing risk on a system marketed as a one-for-one bitcoin representation.

When will Liquid Network restart?

Liquid should restart only after patched nodes agree on one chain, the reserve is reconciled against valid L-BTC liabilities and peg controls can reject implausible withdrawals. Operators have said they are resolving a chain split and preparing a coordinated restart; users have been told not to send bitcoin to peg-in addresses meanwhile.

The recovery improves the balance sheet, not yet the market structure. Returning 85% is materially better than a total loss, but a bridge that can automatically pay nearly its entire reserve against counterfeit claims has failed at its core control point. The cost now falls on L-BTC holders and venues through frozen liquidity and uncertain redemption. A credible restart needs transaction limits, delayed large payouts and independent reserve checks; those changes would trade some speed for the ability to settle without trusting a negotiation after the fact.