This has been slow-boiling since Prime Trust went bankrupt in 2023. If you don't recall exactly what happened, here's a quick summary by Lopp:

Prime Trust was a custodian that acted as the backend for several exchanges and apps. The company was a “qualified custodian” regulated by the State of Nevada.

According to a court filing, Prime Trust migrated its custody onto another platform in 2019. In 2021, the company started unintentionally providing customers with deposit addresses to a 3-of-6 multisig wallet for which it no longer had access to enough keys to sign transactions. Any funds sent to those addresses were lost.

To complete requested withdrawals, the company used customer funds to purchase assets from December 2021 to March 2022. Making matters worse, a crypto bear market set in which placed further strain upon the company’s finances. The company also invested customer funds in TerraUSD, a doomed algorithmic stablecoin that collapsed in May 2022.

By June 2023, rumors began to circulate about the financial hole at Prime Trust. Crypto custodian BitGo agreed to acquire Prime Trust but backed away from the deal. Shortly thereafter, Prime Trust was placed into receivership and eventually filed for bankruptcy.

Okay, so: not your keys, not your coins. We know.

Now from the Blockspace article linked above:

in the case of Swan, Strike, and some others, Prime Trust’s service agreement with these companies created a creditor-debtor relationship, in which the clients were the creditors and Prime Trust was the debtor; further, if this holds true, any withdrawals these companies made in the 90 days before Prime Trust’s bankruptcy should be considered preference-period withdrawals, meaning they are subject to clawback since U.S. bankruptcy code stipulates that a company cannot give any creditor preferential treatment in the three months preceding a bankruptcy.

PCT [the jilted Prime Trust creditors] Litigation Trust heavily leans on the preference-period argument in the Swan and Strike cases, alleging that both companies had insider information of Prime Trust’s imminent bankruptcy (Swan allegedly via direct communication with a Prime Trust executive, and Strike allegedly through meetings with Prime Trust in which it learned of Prime Trust’s regulatory troubles and financial issues).

This alleged insider knowledge is perhaps why the trust is targeting Swan and Strike with such large sums compared to other defendants: each company may have withdrawn the bulk of the assets it held with Prime Trust, while others were making routine withdrawals to keep their businesses functioning.

This is all speculation on my part — nothing in the court documents explicitly says this, but at least in the case of Compass Mining, OpenNode, and Fold, the lawsuit does not claim these companies had insider information.

It seems some of this case comes down to the language in agreements between Prime Trust and Swan or Strike.

If the court does rule that these companies were creditors to Prime Trust, then it would stand to reason that it may also rule that the alleged withdrawals fell within the preference period.

And if the service agreements listed Prime Trust as a fiduciary? Then maybe we would be having a different conversation, but even that’s not a guarantee. Reason being, Prime Trust did not keep separate accounts — either fiat or crypto — for its clients.

As the bankruptcy court’s July 2025 Distribution Opinion put it: “the overwhelming evidence establishes that [Prime Trust] hopelessly commingled assets,” that “the fiat held by the Debtors is not traceable,” and that the “hopeless commingling would not allow the cryptocurrency to be traced.”

So there was no way to clearly establish what belonged to whom, which could have complicated the matter even if Prime Trust had agreements in place to act as a fiduciary and not a contractor.

I'm sure that big holders like Strategy have very clear and indisputable agreements with their custodians, who I'm sure are also very careful not to screw up their multisigs...