Five verbs have carried this whole project: collect, analyze, copy, use, dispose. For seven weeks I’ve built rules for the first four. This is the last one, and it is the verb with the fewest rules of all — which is strange, because it’s the one that decides whether any of the others ever end. Disposition is what happens to your digital property when the government is finally done with it. And the uncomfortable truth is that, right now, the government is almost never done with it.
That’s the thing to see first. Every lifecycle has a beginning everyone watches — the knock at the door, the warrant, the seizure — and an end almost nobody does. We have a century of law about how the government may take your property and almost none about when it must let it go. So the default ending, in practice, is that there is no ending. The forensic image goes into the case file and stays. The copy sits. The data waits. I spent five weeks documenting that inertia — Ganias, Richman, Aboshady, the FBI’s permanent case file, the 702 store queried for years — and § 8 is where the model act finally forces the question the current non-system never asks: now that this is over, what happens to the data?
There are only three possible answers, and a disposition rule is just a statute that makes the government pick one, on a schedule, under supervision, instead of defaulting into permanence. Return it. Delete it. Or keep it. Take them in turn, because each has a trap.
Return. When the case is over, the government gives the property back. Simple for a physical thing — hand back the laptop. But Week 1’s distinction bites here: you can return a device, and you cannot return a copy. Handing someone a copy of their data changes nothing, because the government still has its own. This is exactly what made Richman hard — the court couldn’t “return” the forensic images in any meaningful sense, so returning the data and deleting it turned out to be the same act. For digital property, “return” without deletion is theater. The device goes back; the copies have to go away.
Delete. This is where § 6’s retention clock cashes out. The non-responsive remainder — the intimate 95% the warrant never reached — gets securely destroyed when its clock runs, and the destruction has to be real: not “marked inactive,” not “access-restricted,” but gone, in a way someone can verify. And deletion has to reach every form and every system, which is the § 7 content point made operational. You have not disposed of a wiretap if you destroyed the recording and left the transcript sitting in Sentinel. You have not deleted a seized drive if the image is gone but three derivative extracts live on in a case-management system and a FOIA-reachable file. Disposition follows the content — original, copy, transcript, representation — into every place the government put it, or it isn’t disposition. The one exception is the piece § 6 already built: the sealed reference copy, preserved with a neutral custodian for the genuine appeal-and-innocence interest, and — borrowing Title III’s § 2518(8)(a) — destroyed only on a court’s order, never the seizing agency’s say-so.
Keep. Sometimes the government lawfully keeps the property, because it has been forfeited — the owner’s interest extinguished through legal process. And here is the trap that surprises people: “keep” is not the end of the disposition question. It’s the beginning of a different one. Once the government owns the thing, it still has to decide what to do with it — convert it to cash at auction, or hold it as an asset — and, more importantly, it has to decide who the value belongs to. Because forfeited property is very often not a windfall for the public. It is money stolen from victims. And disposition is not only a question between the government and the person it took the data from. It is also a question between the government and the people that data’s value was stolen from. A disposition statute that forgets the victim is as incomplete as one that forgets the suspect — and forgetting the victim is exactly what happens when “keep” is the default and no rule makes the government account for the value first.
So § 8 does four things. It requires return of the device and of any data the government has no lawful basis to keep. It requires deletion of the non-responsive remainder on § 6’s clock — real, verifiable destruction, reaching every copy, transcript, and derivative in every system — with the single sealed reference copy as the court-controlled exception. For property the government lawfully keeps through forfeiture, it requires rules for conversion and custody, and it puts victim restitution ahead of the public fisc — the government’s claim to forfeited value is junior to the claim of the people it was taken from. And across all of it, it vests the decision to destroy, keep, or convert in a neutral authority, not the agency holding the evidence — because the party that wants to keep something is the wrong party to decide whether it must be let go.
Now the objection, in two parts, because disposition draws fire from two directions.
From the law-enforcement side: deletion destroys evidence you may need later, and forfeiture lawfully funds legitimate operations — tie the government’s hands here and you lose convictions and you starve the agencies that do this work. From a victim-skeptical angle, you’ll even hear that restitution-first is unworkable because victims are hard to identify and their claims are a mess.
I’ve answered the deletion half before and § 8 answers it the same way: the innocence and appeal interests are real, and they are fully protected by preserving the responsive evidence and one sealed, court-controlled copy — not by warehousing everyone’s entire digital life forever on the theory that someday some byte might matter. “Keep it all in case” is the general-warrant instinct; it proves too much.
The forfeiture half is where I want to be precise, because the strongest-sounding objection quietly inverts the right order. Forfeited property only becomes “the public’s” money after the people it was stolen from are made whole. The sequence is not keep-then-maybe-compensate; it is compensate-then-keep-what’s-left. An asset-forfeiture program that treats seized fraud proceeds as a revenue source, with victims as an afterthought to be processed later through some other channel, has the priority backwards — and when the government simultaneously holds those assets as a long-term asset it does not intend to sell, “later” can become “never.” Funding law enforcement from forfeiture is defensible only on the residue, after restitution, under rules. On the money that belongs to victims, the government is a custodian, not an owner, and a custodian’s first duty is to give it back.
That’s § 8, and it’s the keystone that makes the other seven sections durable. Think about what every upstream rule is worth without it. A tight collection rule, a scoped analysis rule, a retention clock, a use restriction — every one of them is a limit on what the government may do with your data while it holds it. But if the holding never ends, every one of those limits is just a speed bump on the road to permanent retention. Disposition is the rule that makes the others mean something, because it is the only one that closes the loop. Collection decides whether the government may start. Disposition decides whether it ever has to stop.
Next Tuesday, § 9: suppression — the first of the enforcement sections, and the question of what actually happens when the government breaks any of these rules. And Thursday, we go full circle — to the most public disposition fight in the country: the battle over the Biden recordings, and a government choosing to publish what it should be returning.
(Model statutory text for § 8 comes in Pass 2 this fall. This post is the why; the drafting is next)
Leave a comment