# Office Policy Reports & Implications for the 2027 Plan

**Date:** 2026-05-24
**Sources:**
- 2017 §1201 Study — *Section 1201 of Title 17: A Report of the Register of Copyrights* (June 2017), at `1201/Policy Studies/Section-1201-of-Title-17-2017/Full-Report.txt` (195 PDF pages).
- 2016 Software Study — *Software-Enabled Consumer Products: A Report of the Register of Copyrights* (December 2016), at `1201/Policy Studies/Software-Enabled-Consumer-Products-2016/Full-Report.txt` (94 PDF pages).

These two reports were the Copyright Office's response to Senators Grassley and Leahy's October 22, 2015 request for policy review. The 2016 study addressed copyright law's application to embedded software broadly; the 2017 study addressed §1201 specifically. Both are policy recommendations to Congress — not regulations and not §1201 rulemaking decisions, but the Office's own analytical positions, which can be cited in any §1201 rulemaking proceeding.

This report synthesizes what the two reports say about repair and what that means for our 2027 multi-class plan (`2026-05-23__2027-Cycle-Multi-Class-Petition-Plan.md`, rendered as `2027-repair-proposals.html`).

---

## Key findings — 2017 §1201 Study

### Finding 1: Permanent repair exemption recommended to Congress

The Office made a **formal recommendation** to Congress for a permanent §1201(a) statutory exemption for diagnosis, maintenance, and repair. The recommendation appears in §III.C.2.b "Obsolescence, Repair, and Modification," PDF pp. 88–96.

Operative language from PDF p. 90:

> "to the extent section 1201 precludes diagnosis, repair, and maintenance activities otherwise permissible under title 17, the Office finds that **a limited and properly‐tailored permanent exemption for those purposes, including circumventing obsolete access controls for continued functioning of a device, would be consistent with the statute's overall policy goals**."

The Office further characterized the triennial rulemaking as a *"somewhat ill‐fitting regulatory approach"* (p. 92, quoting former Register Marybeth Peters' 2001 testimony) for cross-cutting repair issues. The rationale: damaged, malfunctioning, or obsolete TPMs potentially affect *all* classes of works, which paradoxically *defeats* the rulemaking's structural requirement of identifying "a particular class of works."

The 2017 Study endorsed Peters' prior recommendation and went further by extending the recommended permanent exemption to *new* repair activities (not just obsolete-TPM repair). PDF pp. 92–96.

### Finding 2: §117(d) as the preferred statutory anchor

The Office recommended that any repair exemption use the existing §117(d) definitions of "maintenance" and "repair" (PDF p. 94):

> §117(d)(1): "the 'maintenance' of a machine is the servicing of the machine in order to make it work in accordance with its original specifications and any changes to those specifications authorized for that machine"
>
> §117(d)(2): "the 'repair' of a machine is the restoring of the machine to the state of working in accordance with its original specifications and any changes to those specifications authorized for that machine"

The Office's view: "Linking an exemption to section 117(d) would seem to address vagueness concerns, while still providing meaningful relief to many consumers and repair technicians" (p. 94).

This is significant for two reasons. First, §117(d) gives us **statutory text we can mirror** in proposed class language. Second, the "original specifications" framing **pre-rebuts the modification/tinkering objection**: we are asking to restore devices to their original specifications, not to enhance or customize them.

### Finding 3: Recommendation against device-specific limitations

PDF p. 94:

> "the Office recommends against limiting an exemption to specific technologies or devices, such as motor vehicles, as any statutory language would likely be soon outpaced by technology."

This is the Office's **own** position that broader, cross-device repair exemptions are preferable. Cite it whenever opposition argues that a class is "too broad" or "covers too many devices." The Office's stated view is that device-specific class language ages badly.

Implication for the 2027 plan: our broader scopes (Filing 4's full-stack computing, Filing 5's building+energy infrastructure, Filing 2's commercial+scientific pairing) are aligned with the Office's own preferred framing. Don't apologize for breadth — cite the Office's recommendation that breadth is desirable for repair.

### Finding 4: Safety concerns belong outside copyright

PDF p. 94:

> "to the extent that commenters oppose an exemption for repair out of non‐copyright related concerns, such as public safety, the Office believes these matters are better addressed through laws or regulations outside of the Copyright Act."

This is a **direct rebuttal** to the safety arguments commonly raised against repair exemptions in regulated industries (vehicles, medical, energy, building infrastructure). The Office's position: safety is regulated by UL, IEEE, FAA, FDA, EPA, DOE, AHJ inspection, etc. — not by §1201.

Implication for Filing 5: our safety carve-outs (life-safety systems entirely; UL 1741, IEEE 1547 grid-interconnect; EPA/DOE efficiency parameters) implement the Office's own direction. We're not asking the Copyright Office to police safety; we've already carved safety out of scope.

### Finding 5: What the Office rejected

Two narrower proposals were *rejected* in the 2017 Study, and our framing has to respect that:

**(a) §1201(a) infringement-nexus requirement** — PDF pp. 42–46. Some commenters (EFF, Public Knowledge, university libraries) proposed amending §1201(a) so that circumvention is unlawful only if it bears a nexus to copyright infringement (the Chamberlain holding). The Office rejected this because it would "severely weaken the right of copyright owners to exercise meaningful control over the terms of access to their works online." The Office concluded that Congress deliberately decoupled §1201(a) from §106 infringement.

  Our argument cannot be: "TPMs on repair targets are categorically outside §1201 because they don't gate access to copyrightable works." That argument was made by commenters in 2016 and rejected by the Office in 2017.

**(b) Categorical exclusion of "device- or machine-enabling computer programs"** — PDF pp. 47–49. ORI proposed amending §1201(a) to exclude "computer programs that enable the operation of a device or machine." The Office rejected this on line-drawing grounds: operating software in DVD players authenticates copyrighted content, so excluding all device-enabling firmware could weaken anti-piracy protections.

  Our argument cannot be: "device-control firmware is categorically excluded from §1201." We need to keep using the rulemaking, but with the Office's repair-exemption framework as backstop.

### Finding 6: §1201(f) and §1201(j) reform — narrower than we need but supportive

The Office recommended targeted updates to §1201(f) (reverse engineering for interoperability) at PDF pp. 63–71 and §1201(j) (security testing) at pp. 71–84. The §1201(j) recommendations include removing the "sole purpose" language and the multi-factor test ambiguity. Neither directly supports our repair filings, but the broader signal is the same: the Office acknowledges that the existing statutory framework is too narrow for legitimate non-infringing activities and has actively recommended Congress widen it.

---

## Key findings — 2016 Software Study

### Finding 7: §117(a) covers most repair as a matter of underlying copyright

The 2016 Software Study's Repair and Tinkering section (PDF pp. 31–42) found that existing copyright doctrines — particularly §117 — already protect most repair as non-infringing. From p. 35:

> "Properly construed, section 117 'should adequately protect most repair and maintenance activities.'"

The Office's analysis of §117(a) ownership: the device buyer is the §117 owner of the embedded software copy when the software is "invisible to the consumer" — i.e., when it's not the primary purpose of the transaction. From p. 35–36:

> "specialized software controlling certain mechanical components of an automobile, like windshield wipers or transmission, may essentially be invisible to the consumer. In such cases, it would be unusual to characterize the sale of the automobile as involving the licensing of that software."

By contrast, the OS of a personal computer is "a more significant element of the transaction" and is more likely treated as licensed (id.).

This distinction is critical for our 2027 plan:

- **Filings 1, 2, 3, 5 (embedded device-control firmware)** sit firmly on the §117(a) ownership side. Our argument: the device buyer is the §117 owner of the embedded copy and may copy/adapt the program for repair purposes.

- **Filing 4 (general-purpose computing)** sits on the licensing-vs-ownership *line*. Our framing must be: we are asking to circumvent component-authentication chips so the user can replace failed *hardware*; we are not asking to access or modify the *OS*. The 2016 Study's analysis supports this scoping precisely.

### Finding 8: "Repair supports — rather than displaces — the purpose of the embedded programs"

From the 2016 Study at p. 40 (quoted by the 2017 Study at p. 90):

> "the fundamental purpose of any repair is to preserve or restore the functionality of a software‐enabled device so that it may continue to be used. In this respect, repair supports — rather than displaces — the purpose of the embedded programs that control that device."

This is the Office's own framing for our "narrow harm" argument. Repair doesn't compete with the copyrighted firmware — it preserves the firmware's purpose. There is no creative-works substitution, no market displacement, no piracy. Use this quote in every filing's copyright-posture section.

### Finding 9: Opposition's concession in the 2016 record

PDF p. 93 n.508 of the 2017 Study (citing the 2016 record): the joint coalition of AAP, ESA, MPAA & RIAA — typically the largest opponents to §1201 expansions — conceded:

> "automobile repair or other forms of repair might present a unique and distinguishable set of circumstances"

This is a written, public concession from the four largest copyright-owner trade associations that **repair is distinguishable** from the activities §1201 was designed to protect. Use it whenever those same organizations file opposition comments in 2027.

### Finding 10: Office did not recommend statutory change in 2016, but did in 2017

The 2016 Software Study found that existing doctrines (§117, fair use, merger, idea/expression) handle the underlying copyright issues for repair and recommended **no statutory change** at that time. The 2017 §1201 Study went further and **did** recommend a permanent §1201(a) repair exemption.

The reports do not contradict each other: the 2016 Study is about underlying copyright (§106) and concludes existing doctrines suffice; the 2017 Study is about §1201(a) anticircumvention and concludes that a statutory exemption is appropriate because §1201(a) is decoupled from §106 (per the rejection of the nexus requirement in Finding 5(a)). Together they form a complete picture: the underlying activity is non-infringing under §117 and fair use, but §1201(a) over-reaches and warrants a permanent statutory carve-out for repair.

---

## Implications for the 2027 plan

### Filing 1 — Agricultural & Construction Equipment

**Strongly supported by the reports.** The 2017 Study cites repair of "automobiles and farm equipment" (PDF p. 87) as a key example of §1201 overreach. The Office identified Nebraska Farm Bureau testimony specifically (p. 89). Our stationary-ag extension is on solid ground.

**Suggested edit:** add a citation to the Register's permanent-exemption recommendation in the user-community section, and adopt §117(d) language ("original specifications") in the draft class language.

### Filing 2 — Scientific & Laboratory Equipment

**Strongly supported.** This filing was originally framed as a 50/50 commercial-service + scientific pairing; after a strategic reframing (see below), it now stands as Scientific & Laboratory Equipment alone — the commercial-service categories with documented lockouts moved into Filing 5's broader "Commercial Premises Equipment" class. The 2017 Study's permanent-exemption recommendation, the §117(a) ownership analysis (Finding 7), and the 2016 Study's "repair supports — rather than displaces" quote (Finding 8) all apply directly. The 2024 Class 5 deliberate carve-out (Finding 2 in the per-finding list of the 2017 Study, at PDF p. 9 of the Public Knowledge / iFixit Initial Comment) is the §1201-record argument unique to this filing.

**Edits applied:** the copyright-posture entry now cites the 2016 Study at p. 40, the §117(a) analysis at p. 35, the 2017 Study's permanent-exemption recommendation, and the opposition's 2016 concession at p. 93 n.508. The "Why now" section anchors to the deliberate 2024 carve-out and the Office's own permanent-exemption recommendation.

### Filing 3 — American Re-industrialization: Manufacturing Equipment

**Mostly supported, but watch the modification line.** The 2017 Study supports repair but not modification/tinkering for permanent exemption (PDF p. 89, "The Office does not, however, recommend that such a permanent exemption extend to circumvention for purposes of making other lawful modifications to software, or 'tinkering.'"). Our 4 sub-asks are all repair-framed — keep them that way. Don't let the rebranded re-industrialization framing drift toward "factory owners get to modify their equipment."

**Suggested edit:** Filing 3's political-frame section should explicitly note that the filing is **repair-only**, not modification. The "factory owner sovereignty" framing is about *who gets to do the repair* (the operator, not the OEM-only authorized technician), not about *what they can change* (which remains "restoration to original specifications").

### Filing 4 — General-Purpose Computing & Networking

**The doctrinally trickiest filing.** The 2016 Study's distinction between embedded software ("invisible to the consumer," owned per §117) and "a more significant element of the transaction" software (OS, licensed) is directly relevant. Our hardware-only-repair framing already implements this distinction, but it must be explicit and aggressively enforced.

**Edit applied:** the critical-framing section now references the 2016 Study's distinction explicitly. The exclusions (no licensing-bypass, no OS access, no content-DRM defeat) implement the line the Office drew.

**Risk:** opposition will cite the 2016 Study's "OS is a significant element" language back at us. Pre-rebut by emphasizing that our scope sits on the *hardware* side: component-authentication chips, parts-pairing, T2/Secure Enclave hardware verification — not the OS itself. We are circumventing the *gatekeeper between the user and the hardware they own*, not the gatekeeper between the user and a licensed OS.

### Filing 5 — Commercial Premises Equipment

**Strongly supported on every front.** This filing was originally framed as Building & Energy Infrastructure; after the strategic reframing, it now covers the full surface of software-enabled equipment installed on commercial premises — HVAC, BMS, building automation, solar, battery, EV charging, commercial laundry, commercial dishwashers, fitness/wellness, cleaning, hospitality. The diversity of equipment categories is *evidence*, not a list of separate asks — the same TPM pattern across the full surface argues that the class needs to be broad.

Three of the 2017 Study's findings directly underwrite this:
- **Finding 3** (recommendation against device-specific limits, PDF p. 94): the Office's own argument *for* breadth. Cite it as the direct response to any "this class is too broad" objection.
- **Finding 4** (safety belongs outside copyright, PDF p. 94): the Office's own framing for our life-safety / grid-interconnect / EPA-DOE carve-outs.
- **Finding 7** (§117(a) ownership analysis, 2016 Study at pp. 35–36): commercial-premises equipment sits on the "invisible to the consumer" / owned-firmware side of the line — the device buyer is the §117 owner of the embedded copy.

**Edits applied:** the critical-exclusions section cites the 2017 Study at p. 94 directly. The "Why one broad class" section makes the strategic principle explicit: diversity of examples = record depth, not separate asks.

---

## Strategic-level implications

### 1. Reframe the copyright posture across all filings around §117(d) + 2017 Study citation

Our common argument — that the TPMs at issue gate device-control firmware rather than creative works — is now backed by:
- The 2017 §1201 Study's recommended permanent exemption
- The 2016 Software Study's §117(a) ownership analysis
- The 2017 Study's recommendation against device-specific limitations
- The opposition's own 2016 concession that repair is distinguishable

Cite all four. Lead with the Office's *own* position; let opposition argue against the Office's own recommendations.

### 2. The Office's permanent-exemption recommendation is the long-term political backdrop

The 2017 Study recommended Congress adopt a permanent statutory repair exemption keyed to §117(d). Congress has not acted in nearly a decade. That recommendation is still on the table, and every triennial repair win **builds the case** that the rulemaking is doing work Congress should have made permanent.

Use this framing to defuse opposition arguments that:
- The exemption is "too broad" → the Office said breadth is desirable
- The class should "wait for legislation" → the Office said the rulemaking is "ill-fitting" and Congress should act; meanwhile, the rulemaking is our only relief
- The class is "an end-run around the legislative process" → the Office has formally recommended the legislative result we're seeking through the rulemaking interim

### 3. Distinguish repair from tinkering throughout

The Office supports a permanent exemption for repair, maintenance, and diagnosis — but **not** modification or tinkering. Modification stays in the rulemaking process. Our filings are all repair-framed; keep them that way. Don't let political framing (especially Filing 3's re-industrialization branding) drift toward "owner sovereignty to modify."

The §117(d) "original specifications" language is our doctrinal guardrail.

### 4. Limits of what the reports support

The Office did *not* endorse:
- A categorical §1201(a) nexus requirement (Finding 5(a))
- A categorical exclusion of device-control firmware from §1201(a) (Finding 5(b))
- Permanent exemption for modification/tinkering

Our framing must respect these limits. We cannot argue our TPMs are categorically outside §1201(a)'s scope, and we must keep our scope on repair, not modification.

### 5. "Broad class" as pretext for "thin record" — the master strategic principle

Reading the Office's denial history across cycles, the "non-cohesive class" framing repeatedly co-occurs with a thin evidentiary record. Where the record is concrete and named — vehicles (2015), consumer devices (2018), medical devices (2021), food-prep (2024) — the Office accepts substantial breadth. Where the record is thin or speculative — broader commercial / industrial in 2021 (Class 12) and 2024 (Class 5 broader scope) — the Office reaches for "non-cohesive class" as the denial rationale.

The actionable principle: **counter the broad-class objection by building the record, not by shrinking scope.** Specific examples in a filing are record-depth contributions, not arguments for separate filings. The diversity of examples *is the argument* that the class needs to be broad. This is exactly the framing the 2017 Study supports at PDF p. 94 ("recommends against limiting an exemption to specific technologies or devices").

This principle drove the 2027-plan restructuring: Filing 2 became Scientific & Laboratory Equipment alone (clean record advantage on the deliberate 2024 carve-out), and Filing 5 absorbed the commercial-service categories under a broad "Commercial Premises Equipment" frame. The disparate equipment categories in Filing 5 are not asks for separate paragraphs — they are evidence of a systemic problem that one broad class is the appropriate response to.

---

## Citations index

| Topic | 2017 §1201 Study | 2016 Software Study |
|---|---|---|
| Permanent repair exemption recommended | PDF pp. 88–96 | — |
| §117(d) as statutory anchor | p. 94 | — |
| Recommendation against device-specific limits | p. 94 | — |
| Safety belongs outside copyright | p. 94 | — |
| §117(a) ownership analysis | — | pp. 35–36 |
| "Repair supports rather than displaces" | p. 90 (quoting Software Study) | p. 40 |
| Opposition's concession on repair | p. 93 n.508 | (2016 record cited there) |
| Nexus requirement rejected | pp. 42–46 | — |
| Device-enabling exclusion rejected | pp. 47–49 | — |
| Rulemaking "ill-fitting" | p. 92 (quoting Peters 2001) | — |
| §1201(f) reverse engineering reform | pp. 63–71 | — |
| §1201(j) security testing reform | pp. 71–84 | — |
| Repair vs. modification distinction | p. 89 | (less prominent) |

All page numbers refer to `[PAGE N]` markers in the `.txt` extractions (= PDF page numbers).

---

## Not legal advice

This document is a research synthesis of two Copyright Office policy reports. It is not a legal conclusion and does not constitute legal advice. The reports are the Register's recommendations to Congress — they are not regulations and do not bind the Office in any rulemaking proceeding. A petitioner should work with counsel to verify each citation and to assess its persuasive weight against the specific record being built in 2027.
