Demand The Full Campus Camera Inventory And The Retention Record

:::caution Legal Disclaimer Tyler Robinson is charged, not convicted, and is presumed innocent. Nothing on this page states as fact that any living person committed a crime, destroyed evidence, or acted improperly. Court filings in this case are subject to a gag order and sealing, so the public record is incomplete by design. :::

The One Thing We Would Do Differently

Serve a Rule 14 subpoena duces tecum on Utah Valley University for the complete campus camera inventory, the video-management-system configuration and per-device retention settings, and the machine-generated export log — then set that log against what the State actually produced.

This is the campus half of the withheld-material thesis, and it should be kept rigidly apart from the federal half. Nothing here needs an intelligence agency to answer anything. The custodian is a Utah public university sitting eight miles from the courthouse, the records are ordinary institutional records, and the process is a state subpoena that a judge signs without ceremony. As a motion this is close to a certainty. The federal half — classified holdings, agency decisions that reached the courtroom secondhand — depends on a gateway ruling and is a different animal entirely. Do not let the two be argued together, because the hard one drags down the easy one.

The insight the demand is built on is that the State produced campus footage. On the State's own 5-6-26 preliminary-hearing exhibit list, Exhibit 12 is "UVU surveillance 9/10/2025" at Bates 1448-1463 — sixteen Bates numbers. Exhibit 13 is "Doorbell camera surveillance 9/10/2025" at Bates 15215-15226. Four sub-exhibits from the Exhibit 12 family — 12.1, 12.2, 12.3 and 12.4 — were admitted on Day 4, 9 July 2026, and the certified transcript records the Court distinguishing 12.4, "the unmarked, non-zoomed version," from 12.1, the enhanced version, which counsel described as running "eight, give or take two, minutes."

So footage exists and footage was produced. A defense therefore never asks did you produce footage. It asks: produced from how many cameras, out of how many that exist? Sixteen Bates numbers is a numerator. Without the denominator it means nothing at all — it could be complete coverage or it could be sixteen items out of four hundred. And nothing on the list we hold even says what those sixteen numbers are: sixteen camera views, sixteen exported clips, or sixteen still frames pulled from one clip. Exhibit 13 has the same shape and the same hole — twelve Bates numbers for "Doorbell camera surveillance," with no statement of how many doorbells were canvassed, how many were recording, and how many declined.

The demand goes by subpoena to the university, not by a discovery request to the prosecutor, and that choice is deliberate. On 2 July 2026 the State answered a defense request for law-enforcement witness records by saying, among other things, that "the Utah County Attorney's Office does not possess these materials," and that Rule 16(a)(4) good cause did not justify compelling "the production (and perhaps creation) of the records." Whatever one makes of that position, it forecasts the answer to any Rule 16 request for a third party's system configuration. A prosecutor cannot produce a university's device table because a prosecutor does not have one. Rule 14 goes around the argument instead of into it.

There is also a route that needs no judge at all. Utah Valley University is a public institution and its records are reachable under the Government Records Access and Management Act, Utah Code Title 63G, Chapter 2. A defense investigator can request the camera inventory and the retention schedule as public records, and a GRAMA denial is itself a dated, written document from the custodian saying what it holds and why it will not give it up. Either the records arrive cheaply or the refusal becomes an exhibit to the subpoena motion. There is no version of sending the request that leaves the defense worse off.

The rider we would attach to the subpoena, in this order:

The complete camera inventory for the campus as configured on 10 September 2025: every device, by unique identifier and by the name it carries inside the system, with its physical location, mounting height, field of view or coverage description, sensor resolution, frame rate, whether it records continuously or on motion, and its individual retention setting.
The system map and the VMS configuration export — which cameras were bound to which recorder, which server, and which storage volume, and any grouping or "view" definitions that determine what an operator sees together on one screen.
The retention schedule and the actual overwrite behaviour, per device. The policy on paper and the behaviour in fact are different records and both are demanded. Storage fills unevenly, so a nominal thirty-day policy can be a nine-day reality on a high-bitrate camera.
The export log — every clip pulled out of the system: by which user account, on what calendar date and clock time, covering which camera, spanning which interval, in which format, and to what destination.
The litigation-hold or preservation notice, if one issued: its text, its date, who issued it, who received it, and which devices or date ranges it covered.
The health and outage log for the operational period: which cameras were offline, degraded, misaimed, or recording to a failed volume, and since when.
A proposed protective order tendered with the subpoena — attorneys'-eyes-only treatment, camera locations by zone rather than street address — because a campus camera map is a genuine security record and offering the protection first is what gets the document produced instead of quashed.
The production log the State can be made to set against all of the above — a device-by-device statement of what was collected, what was produced in discovery, and at which Bates number.

What The Defense Team Did Instead

The public record shows real preservation work, early, and it should be said plainly. On 8 October 2025 the defense filed a Motion to Preserve Evidence, docket 75. The Court granted it on 20 October 2025, docket 107. The State moved to set that order aside on 22 October 2025 as prematurely entered, docket 116, and on 4 November 2025 Judge Graf entered a stipulated order, docket 161, setting aside the first order and granting the motion in part. The 30 September 2025 Request for Discovery had already asked, at item 12, for all photographs and "video tapes pertaining to this case in the possession of any law enforcement or government agency."

The stipulated preservation order is the document worth reading closely, because of what it reaches. On its face it orders "the Utah County Attorney's Office, its agents and any law enforcement agencies involved in the investigation of this case" to preserve evidence in their custody or control. That is the prosecution side of the ledger. A third-party institutional custodian's own recording system is not obviously inside those words, and an order that does not name a custodian is an order that custodian's IT department will never see.

What is not visible in the archive we hold is any subpoena duces tecum directed to the university for the inventory, the retention settings or the export log. That is a statement about our material and nothing more. The archive holds 374 PDFs against 867 known docket entries, only 258 of which ever had a public PDF, and the filings in this case are gagged and partly sealed.

The authentication line, by contrast, was plainly pursued and preserved. On 3 July 2026, three days before the preliminary hearing opened, the defense filed Standing Objections to the State's Exhibits, docket 238, objecting under Utah Rule of Evidence 1102(b)(3) and article I, section 12 of the Utah Constitution to the use of hearsay "to lay foundation and authenticate all of the State's exhibits including photographs and video recordings for which the attesting officer does not have personal knowledge." The certified Day 4 transcript then shows the Court noting that standing objection as each exhibit in turn was admitted. Whether the officer describing the campus video had personal knowledge of how it was recorded and exported is a question the defense put on the record and kept there. What this page adds is not that objection but the documents that would give it teeth.

Two further public facts cut the other way and belong here. The 28 August 2026 Discovery Status Report records 45 productions and 29,237 files, and expressly lists "surveillance video from public and private entities and persons" as a produced category — so video from third parties is flowing. And on Day 4, Sergeant Jennifer Faumuina testified that the investigation went back for a better copy: the command-post video "was pretty grainy at the time," and "it wasn't until we were actually able to get the footage directly from UVU" that the image could be zoomed. Somebody re-exported. That is diligence, and it is also, exactly, an export event that a log recorded.

Where This Belongs In The Case

Stage of the case	When that stage runs	Status of this step
Third-party subpoena duces tecum, Utah R. Crim. P. 14	Available from arraignment; available now	Not visible on the public record
Preservation motion practice	Filed 8 October 2025; stipulated order 4 November 2025	Done — scope reaches the prosecution side
Preliminary hearing cross	6-10 July 2026 — completed	Exhibit 12 family admitted; inventory not on the record we hold
Spoliation motion and adverse-inference instruction	After bind-over, before trial	Not yet due — and a step not yet due has not been missed

Witness Or Discovery Target

How it would be done	Target	Took the stand?	Status on the public record
Rule 14 subpoena duces tecum, never served publicly	UVU records custodian and the video-management-system administrator	No	No public subpoena or return
Cross-examination — witness testified, question not asked	Jennifer Faumuina, SBI sergeant, the case's "traffic manager"	Yes — 7 July and 9 July 2026	T1. Certified Day 4 transcript searched: she describes a grainy command-post copy and a later copy obtained "directly from UVU". No question appears about how many cameras exist
Cross-examination — foundation witness for the campus video	David Hull, sergeant	Yes — 6 and 7 July 2026	T3 only. Days 1 and 2 are not certified; we hold machine transcriptions of a pool feed, which are not the record. Nothing about his testimony can support a "not asked" claim
Records demand against the State	Device-by-device production log set against the inventory	n/a	Not visible in the archive we hold
Witness never called	UVU campus-safety or IT officer able to authenticate the export log	No	Never subpoenaed publicly

The Questions That Were Not Asked

Addressed to Sergeant Jennifer Faumuina, who testified on Day 4 and whose certified transcript we have searched:

Sergeant, how many cameras were operating on the Utah Valley University campus on 10 September 2025?
Did anyone on your team ever obtain a list of them?
You told us the copy at the command post was grainy, and that you later got footage directly from UVU. On what date did that second acquisition happen?
Who at the university performed that export, and what did you ask them for — a camera, a building, a time window, or everything?
When you say the suspect was "out of camera's view" after crossing toward Campus Drive, is that your conclusion from the clips you were given, or from a map of where the cameras actually point?
Did anyone ask the university what its recording-retention period is?
Was any written preservation or hold notice sent to the university, by anyone, at any time?

Addressed to the university's records custodian and system administrator, if produced under subpoena:

How many cameras were configured in the system on 10 September 2025, and how many of those were recording?
What is the retention period for each of them, and does the system delete by age, or by overwriting the oldest data when a volume fills?
Does your video management system write an audit entry each time a user exports a clip? Genetec, Milestone, Avigilon, Verkada and Axis all do. Which product is this, and where does that log live?
Please read the earliest export entry for 10 September 2025 and tell the court the user account, the time, the camera, and the interval exported.
How many distinct cameras appear anywhere in that export log for the period 10 September to 10 October 2025?
Did the university issue a litigation hold? On what date, by whom, and covering which devices?
Which cameras were offline, degraded, or recording to a failed volume that week?
For any camera that was neither exported nor held, when did its recording of 10 September 2025 cease to exist?
Was the doorbell canvass logged the same way — how many residences were approached, how many cameras were recording, and how many owners declined?

Why A Defense Attorney Would Want This

The export log is the hinge, and everything above is scaffolding around it. A video management system is a database with an audit trail. It writes a row every time a human being pulls a clip out of it: who, when, which camera, which minutes. Nobody has to remember anything and nobody has to admit anything. The log is machine-generated, contemporaneous, and indifferent to the outcome of this case.

Two subtractions follow, and they are the whole point:

Inventory minus export log equals the footage that existed and was never taken. Every camera that was recording that day and appears nowhere in the export log recorded something that no investigator ever looked at.
Export log minus production log equals the footage that was taken and was not produced. Every clip pulled from the system that carries no Bates number is a clip the defense has not seen.

Neither number is an accusation. Both are arithmetic, and either one is a closing argument that a juror can follow on their fingers.

Then the retention question turns arithmetic into a remedy. A duty to preserve attaches when litigation is reasonably anticipated, and in a homicide on a university campus that is the same afternoon. If the system's own settings show that recordings from 10 September 2025 overwrote after that date, on hardware the institution controlled, then material that a preservation notice would have saved is simply gone. Utah is friendlier ground for that argument than the federal courts: the Utah Supreme Court's lost-evidence decision, State v. Tiedemann, 2007 UT 49, declined to import the federal bad-faith requirement of Arizona v. Youngblood into the state due-process analysis and instead directed courts to weigh the reasonableness of the conduct, the prejudice to the defendant, and what remedy fits. An adverse-inference instruction is squarely within that menu. That is a real, grantable trial remedy — a sentence the judge reads to the jury — and framing the motion around the remedy rather than around blame is what makes it grantable.

Retention is not a hypothetical failure mode in this case. Commentary compiled in the investigation file describes a news organisation requesting the Washington County footage of the 11 September 2025 surrender, being kept waiting, and then being told the recording had passed a thirty-day retention window and no longer existed. That account is T3 — secondary commentary, not a filing and not a certified transcript, and it is about a different custodian at the other end of the state. It is offered only for the shape of the thing: a public-agency recording system, a request that arrives after the cycle has run, and an answer that is neither a denial nor a production. The State's own exhibit list includes Exhibit 15, "Video at WCSO 9/11/2025" at Bates 20853, so some video from that building was collected and produced. Which is the point exactly — footage that somebody exported in time survives, and footage nobody exported does not, and the export log is the document that says which was which.

It also feeds the identity fight without needing a theory. The government's account tracks a figure across campus from roughly 11:50 a.m. Every gap in that track is a place where the person leaving one frame and the person entering the next are an inference. A camera map tells a jury whether the gaps are inherent in the coverage or whether a camera was pointed at the gap and its footage was never pulled. And it is the only document that tests the sentence "the suspect was out of camera's view after that."

And the inventory is the only document that says what was pointed at the canopy. Every frame of campus video that has surfaced in this case runs on the government's chosen axis: the approach across the grass, the tunnel, the Losee stairwell, the roof, the drop, the run north. That is a track of a person. It is not a picture of what happened to Charlie Kirk. This site's evidence review concludes he was killed by an explosive device, most plausibly a shaped charge in or on the microphone, and the State's rifle account is the contested proposition — so the frames a defense wants most are the ones aimed at the courtyard, the canopy and the lectern, at the moment of the fatal wound, from an institutional camera with a fixed mount and an embedded clock. A camera map answers, in one page, whether any such camera existed. If one did and its footage was never exported, that is the single most consequential line in the whole document. If none did, that is worth knowing too, and it is worth knowing from a records custodian rather than from an argument.

This page serves Route 2 — the investigation itself. It needs no alternative perpetrator and it does not depend on winning the mechanism fight first. The closing sentence it builds toward is short: "They gave you sixteen items of campus video. They never told you how many cameras there were. Their own system wrote down every clip anybody took out of it, and nobody in this courtroom has shown you that list."

Day-one date: a written preservation and litigation-hold letter to Utah Valley University on 11 September 2025, the day after, delivered to the general counsel and to campus IT, followed by an ex parte motion for an order to preserve if it went unanswered inside seven days. On that date every camera on campus still held its 10 September recordings, whatever the retention period was, because no plausible cycle is shorter than a day. On a thirty-day cycle the unexported cameras erased themselves around 10 October 2025; on the fourteen-day cycle that busy institutional systems often fall back to, around 24 September 2025. The Motion to Preserve Evidence reached the docket on 8 October 2025 and the stipulated order landed on 4 November 2025 — both after the shorter of those windows had closed, and the order does not on its face name the university. This is exactly the perishability point: the letter is free, it takes twenty minutes, and it is the difference between a demand for documents and a demand for documents that still exist.

The Honest Counterargument

The strongest answer is the dullest one, and it is probably correct: universities overwrite footage constantly, and that is storage economics, not concealment. Campus systems run hundreds of cameras against finite disk, and the oldest data is destroyed automatically by design, on a schedule nobody consults and no human being executes. A recording that vanished on 24 September 2025 vanished because a disk filled, and Tiedemann still asks about the reasonableness of the conduct — routine automated overwriting by a non-party is close to the definition of reasonable. A judge may well say so.

The second answer is Faumuina's own testimony, and it cuts against us. The investigation was not content with the grainy command-post copy; somebody went back to the university and obtained a better one, good enough to zoom in and resolve whether a hand touched a window. That is a record of careful collection, not of indifference, and the State will read those lines to the court before we finish reading ours.

Third, an inventory can be a trap. If the subpoena returns four hundred cameras and three hundred and eighty of them point at parking meters, loading docks and vending machines, the defense has spent its credibility to produce a witness who will explain, patiently and persuasively, why coverage of the relevant sightlines was always going to be thin. A denominator only helps if the denominator is embarrassing, and we do not know that it is.

Fourth, the university has a real and respectable objection waiting, and it is not obstruction. A complete map of where a campus's cameras point, what they cannot see, and how long they remember it is a security document. Utah's own records statute anticipates this: GRAMA classifies as protected those records whose release would jeopardise the security of government property or systems, at Utah Code section 63G-2-305, and a public university would be within its rights to invoke it and to move to quash a subpoena on the same ground. The answer is not to pretend the objection is pretextual. It is to ask for the material under a protective order, attorneys'-eyes-only, with locations described by zone rather than by address, and to accept a redacted inventory that still carries the count, the retention setting and the export history. A defense that demands the whole thing in open court will get nothing; one that offers the protective order first will usually get most of it.

Fifth, and most likely of all: counsel of record may already hold every one of these documents. Twenty-nine thousand two hundred and thirty-seven files across forty-five productions is a great deal of paper, the produced categories expressly include third-party surveillance video, and an export log is a small file that would attract no attention on its way past. The absence of a public subpoena is at least as consistent with they already have it as with anything else.

Sources On This Site

:::danger Theoretical Exercise Only This page is a theoretical exercise. It is a written simulation of how an idealised defense attorney might approach this case. We make no claim that there is any problem with the current defense attorney, with any member of the defense team, or with any decision they have made.

Real capital defense is conducted under a gag order, with sealed filings, and on the basis of full discovery that the public has never seen. Counsel of record know facts we do not. A step described here as "not on the public record" may already have been taken privately, may have been considered and rejected for good reason, or may not yet be due at this stage of the case. Silence in the public record is not evidence of a failure.

Nothing here is legal advice, a criticism of any licensed attorney, or an allegation of professional misconduct. :::

This website's author is @HolonCitizen on X and YouTube/@HolonCitizen. Follow me there.

The One Thing We Would Do Differently​

What The Defense Team Did Instead​

Where This Belongs In The Case​

Witness Or Discovery Target​

The Questions That Were Not Asked​

Why A Defense Attorney Would Want This​

The Honest Counterargument​

Sources On This Site​