Liquid's exit stays shut as Blockstream refuses the ransom

Blockstream's Sept. 11 answer to the Liquid ransom demand landed three days after Liquid's Sept. 8 accounting showed 598.5 bitcoin from the Sept. 6 exploit still outstanding, and it paired a refusal to pay with an outright rejection of the takers' white-hat framing. "Taking assets without authorization and withholding their return is a crime, not responsible disclosure," the company said. "It is not white-hat activity. It is theft."

That refusal is easy to defend and costs the company little; the harder count sits beside it. Peg-outs, the only mechanism that moves bitcoin out of Liquid, remained disabled even after block production and transactions came back Sept. 10. The exit stayed shut.

The exploit never touched a private key, per Liquid's Sept. 8 incident report; it worked through the way nodes running the open-source Elements software cached range-proof verifications, allowing roughly 4,000 L-BTC with no bitcoin behind it in the network's reserve to be accepted as valid and pushed through Liquid's standard peg-out process to release about 4,000 BTC. On Sept. 7, 3,400 BTC came back to the Liquid Federation's peg wallet; by the Sept. 8 accounting, 598.5 BTC, or 15% of the total, had not, and the actors had left a message on the Bitcoin blockchain identifying themselves as white-hat security researchers asking to be contacted about the vulnerability.

Those figures reconcile to a single event—3,400 returned plus 598.5 outstanding is 3,998.5 against roughly 4,000 released—and Liquid's response was an emergency Elements release, v23.3.4, shipped Sept. 9 to address the proof-verification cache vulnerability, with peg-outs kept off as a precaution pending what the network calls the final stage of recovery.

As this publication has argued, federated custody is a control illusion, and this episode is about as clean a demonstration of it as the model will produce. The recovery that worked was a negotiation conducted onchain with the address that drained the network, rather than the exercise of a control point; the 15% left behind is what a model with no named party behind the freeze costs its users. Blockstream's own argument—that paying would leave developers of open-source Bitcoin software facing ransom demands exceeding their economic participation—is correct, and it points further than the company takes it: a chain whose holders cannot withdraw without federation consent is custody in all but name, and a qualified custodian at minimum names who holds the keys and who answers for the loss.

Until peg-outs return, the figure to watch is 598.5—the bitcoin still outstanding after a negotiation the federation's own accounting closed 15% short.