Owning data is not permission to sell it

Three words get used interchangeably in estate discussions of data, and they mean different things. Possession is having the records. Ownership is holding title to them. Permission is the right to transfer them to a particular buyer for a particular use. A sale requires all three. The third is the one nobody has usually checked, because the estate simply stepped into whatever position the debtor built over the years.

That position is contractual, and it survives the filing. The privacy policy that was live when a customer signed up. The consent language in a clinical study. The vendor agreement that let a partner's records flow into the warehouse. Bankruptcy is powerful, but it does not dissolve these obligations. They follow the data into the sale, and they bind the buyer after it.

The 23andMe case put this on the front page. The court appointed a consumer privacy ombudsman, customer consent rights shaped what could transfer, and the approved sale required the buyer to honor the company's existing privacy commitments. The most valuable asset in the case was sold inside a fence built years earlier by product lawyers who never imagined a bankruptcy. That is not the exception. Every consumer-facing debtor has such a fence, and the estate inherits it as is.

The common mistake is sequencing. Treated as a closing condition, a rights problem surfaces after months of estate-funded work and kills the sale, and the professionals' time is an administrative expense either way. Surfaced in the first weeks, the same problem usually has an engineering answer: de-identify the records, filter the affected cohort, license only the uses the consents cover. The difference between a dead deal and a narrower one is rarely the problem itself. It is when the problem was found.

There is also a price effect that runs against intuition. Laboratories do not pay less for a dataset with documented rights limitations. They pay more for it than for an identical dataset with undocumented ones, because what they are buying is the absence of future litigation, and a well-drawn fence is the proof. A rights memo is not compliance overhead. It is packaging, and it is one of the few forms of value a trustee can add to this asset for the cost of a few conversations with counsel.

The rule, then: in the first weeks of the case, with counsel in the room, ask what may be sold, to whom, and for what use. It is a smaller question than it sounds, and every decision that follows gets easier once it has an answer.