Who Gets to Say Yes?

Two weeks in, the model digital property act has an object and a default. Week 1 named the thing being protected — the device, the data, the copy — and insisted that “papers” means content. Week 2 set the default rule for that thing to no: the government may not collect, analyze, copy, use, or dispose of digital property except as the Act allows. A default of no is only useful if there’s a disciplined way to get to yes. That’s this section. Authorization to collect is the first and most important permission the statute grants, and the question it answers is deceptively simple: who gets to say yes, and what do they have to show?

Title III answers that question with unusual seriousness, and it’s worth seeing how many gates it builds. To wiretap someone, the government must pass through two separate doors. The first is inside the executive branch. Under 18 U.S.C. § 2516, an application for a wiretap can’t be made on the say-so of the agent who wants it, or even the local prosecutor. It has to be personally authorized by a named, high-ranking official at Main Justice — the Attorney General, a Deputy or Associate, or a specifically designated Assistant Attorney General. Congress decided that intercepting private communications was so intrusive that someone near the top of the Department had to put their own name on the request before a court ever saw it. The second door is judicial: under § 2518 a federal judge has to find probable cause, has to find that the surveillance is properly limited, and — this is the part people forget — has to find necessity, meaning the government has shown that normal investigative techniques have been tried and failed, or would be futile or too dangerous. You don’t get a wiretap just because one would be useful. You get it because you’ve shown you need it.

Stack those up and you see the design. High internal approval concentrates accountability in a senior official who can be identified and answer for the decision. Judicial probable cause is the constitutional floor. Necessity forces the government to justify the intrusiveness of the method, not just the existence of a crime. Three requirements, each doing different work, all triggered by one judgment: that this particular technique is invasive enough to deserve the friction.

Now hold that up against how the government gets authorization to collect digital property today, and watch the whole structure invert.

To image an entire smartphone — a device the Supreme Court has told us holds “the privacies of life,” more revealing than a month of intercepted calls — the government generally needs a warrant. That’s real, and it’s the judicial door, and it matters. But the other requirements mostly vanish. There is no § 2516 analog: no rule that a senior official has to personally authorize the comprehensive imaging of a person’s digital life. There is usually no necessity requirement: the government does not have to show that a narrower search wouldn’t do, that it tried less-intrusive methods first, or that it needs the whole device rather than a slice of it. And the particularity that the warrant nominally requires tends to evaporate at the collection stage, because the government takes a complete forensic image regardless of how narrow the warrant’s stated object is. Then step below the warrant line and it gets starker: enormous quantities of digital data are acquired with no warrant at all — subpoenas, court orders on less than probable cause, consent, and outright purchase from data brokers.

Put plainly: the authorization required to intercept a handful of a suspect’s phone calls is, by design, harder to obtain than the authorization required to copy that suspect’s entire digital existence. The formalities runs inversely to the intrusion. That is not a considered policy choice anyone made. It’s an accident of the fact that we wrote a careful statute for the telephone and never wrote one for the drive.

So § 3 of the model act does for collection what § 2516 and § 2518 did for wiretaps, with one update: it makes the authorization scale with the intrusiveness of what’s being collected. There are two knobs to turn, and the section turns both together. The first is the predicate — what the government must show: probable cause always, particularity that actually binds the scope of collection, and, for comprehensive collection, a necessity showing borrowed straight from Title III. If you want to image the whole device rather than extract a defined category of data, you should have to justify why the narrower method won’t serve. The second knob is the approval level — who inside the government has to sign before going to the magistrate. Routine, targeted collection can run on an ordinary warrant and ordinary internal approval. But the most intrusive asks — imaging an entire digital life, bulk or programmatic collection, collection that will feed years of retention and re-analysis — should require elevated, named authorization the way a wiretap does, so that a senior official is accountable for the decision to sweep everything.

Here’s the strongest objection, and it’s a practical one. Title III’s authorization process is notoriously slow and cumbersome — so much so that agencies often route around it toward authorities with lighter requirements. Impose a § 2516-style approval requirement and a necessity showing on device searches, the argument goes, and you either grind ordinary investigations to a halt or you invite endless litigation over whether “normal techniques” were really exhausted. Digital evidence is also perishable — accounts get deleted, devices get wiped — so the time a heavyweight authorization process consumes is time in which the evidence disappears. And the sheer volume of device searches dwarfs the number of wiretaps; what’s workable for a few thousand intercepts a year may be unworkable for the routine phone on every arrestee’s person.

The volume point is the serious one, and it’s exactly why § 3 tiers rather than imposing one heavy standard across the board. Not every collection needs the full apparatus. A targeted extraction under a narrow warrant should be easy; that’s most collection, and the friction there should be light. The heightened predicate and elevated approval attach only to comprehensive collection — the whole-device image, the bulk pull, the acquisition designed to be retained and re-searched. That is a much smaller set of cases, and it is precisely the set where the intrusion is greatest and where, today, the ceremony is lowest. Tiering the authorization to the intrusion is administrable — Title III already tiers, and so does the Fourth Amendment in a dozen other contexts — and it directly attacks the “image everything because it’s easier” reflex that produces the retention problems this blog keeps documenting. As for perishability: that’s an argument for a fast, well-designed emergency track with after-the-fact review, which the Act will build, not an argument for having no elevated authorization at all.

One more thing § 3 makes visible. A warrant and an internal authorization are not the same safeguard, and we’ve quietly let the first stand in for the second. The magistrate checks the government’s showing against the Constitution. The internal approval level decides who in the executive branch owns the decision — whose name is on the choice to reach into someone’s entire life rather than a corner of it. Title III insisted on both because they guard against different failures. A digital collection statute should too.

Name the object. Set the default to no. Then make the first yes a calibrated one — harder to obtain the more of a person’s life it reaches. Next Tuesday, § 4: once you’re authorized to collect, what the procedure has to look like — particularity that bites, minimization/scoping at the moment of collection, and the record that proves what was taken. And Thursday, we jump to the very far end of the lifecycle, into the strangest asset the government now seizes, to watch what happens when there are no disposition rules at all.

(Model statutory text for § 3 comes in Pass 2 this fall. This post is the why; the drafting is next.)