The institution with the real choice is the North American Electric Reliability Corporation (NERC), and the deadline is December 31, 2026. On July 16, 2026, the Federal Energy Regulatory Commission (FERC) ordered NERC, under section 215(d)(5) of the Federal Power Act, to develop new or modified Reliability Standards addressing the reliability risks of computational-load integration into the Bulk-Power System, and to revise its Rules of Procedure to include registry criteria for computational-load entities. Both deliverables are due by the end of the year. NERC's immediate choice is how to propose registry criteria and standards by the deadline; those proposals will determine whether and how computational-load entities later become accountable actors in bulk-power reliability governance.
This brief is the institutional angle on the reliability track. It is the companion to Juncture's live large-load tracker at /ferc-large-load-governance/, which covers the tariff and interconnection track. The two tracks are separate instruments in the same policy project, and they should be read together, not as duplicates.
What changed
The order, issued in Docket RD26-7-000, is a section 215(d)(5) directive: the provision under which FERC directs NERC to develop a Reliability Standard and submit it to the Commission. The FERC July meeting summary, opened this run, states that the order directs NERC to develop standards "addressing the reliability risks associated with the integration of computational loads into the Bulk-Power System" and to revise its Rules of Procedure "including registry criteria for computational load entities."
Two implications follow. First, the order creates the premise that computational load is a distinct reliability category. The order directs NERC to develop registry criteria for computational-load entities. It does not itself create a registry, set a threshold, or establish that any particular data center must register, report, or comply. Those questions remain prospective until NERC proposes criteria and FERC acts on them.
Second, the registry and the standards are two different instruments that must move in parallel. The registry defines who is in the system. The standards define what they must do. The order directs NERC to develop both by December 31, 2026, which makes the deadline a single point of accountability for both definitions.
The order is also the companion instrument to the June 18, 2026 large-load tariff orders that FERC issued to six regional transmission organizations (RTOs) and independent system operators (ISOs). Those orders addressed who pays and who connects. The computational-load order addresses how these loads are governed as reliability actors. The December 31, 2026 deadline gives NERC and stakeholders roughly five and a half months to define a regulatory category that did not exist when the year began.
The correction
The lazy consensus reads the order as a routine delegation to the standards body, of the kind FERC issues regularly, and therefore as a process event with no content. The correction is that the order creates the category before the category has any content, and the content is where the policy is made. The registry criteria do not exist at this writing, and no one can state them as fact. Three open design questions will determine the practical reach of the new category.
First, registry thresholds. At what size does a computational load become a registered entity? A threshold measured in megawatts, in demand profile, or in interconnection status will define how many data centers are captured. The thresholds have not been proposed, and nothing in this brief describes them as if they exist.
Second, the reliability risk model. The standards must translate "computational-load integration" into specific requirements. The candidate risks include rapid demand swings, co-located generation interactions, and the operational profile of AI workloads, but the standard content is prospective and not yet drafted.
Third, the interaction with the tariff track. The computational-load standards will interact with the six operators' large-load tariff responses under the June 18 orders; how registry obligations align with cost allocation and interconnection reform is unresolved.
Under Juncture's rules of evidence, the proposed NERC rules are kept prospective. This brief makes no claim about standards or thresholds that have not been proposed.
Why this matters now
The decision window is the second half of 2026, bounded by a fixed date. NERC must propose thresholds and standards that reconcile two tensions: registration burdens proportionate to reliability risk, and a category that cannot be evaded by being small or served off the bulk-power system. How NERC resolves those tensions will define the category for years, because the first registry criteria become the baseline for every later amendment. Data-center operators, co-location providers, utilities, and transmission owners all have an interest in the shape of the registry, and the stakeholder filings in the docket are where that interest becomes visible. The September 7, 2026 comment deadline on the Department of Energy's National Transmission Needs Study is a related, separate point on the same calendar.
The institutional constraint
The constraint is that NERC must produce both deliverables on a compressed calendar, and its product then passes through FERC review, which can remand or reject. NERC's standards-development process normally runs on a multi-year cycle with stakeholder balloting; the December 31 deadline compresses that cycle for a category with no precedent. NERC must also reconcile the registry with existing registration categories and avoid creating gaps that let loads evade obligations by being small or served off the bulk-power system. The deadline is a single point of accountability, both a forcing mechanism and a quality risk: a submission made on time but drafted in haste would hand FERC a review problem instead of a standard. FERC's review posture, strict or permissive, will determine whether the category gains real content on the first pass or needs a second round.
What the consensus misses
The consensus treats reliability governance as a technical backwater that will absorb data centers without changing the industry's shape. The correction is that bringing hyperscale load growth inside governance frameworks designed for a different electricity system is a structural change, not an administrative one. The June 18 orders and the July 16 order are separate instruments, but they are the same policy project: converting the largest new load class in a generation into a governed actor class with registration, reporting, and standards obligations. The consensus also misses that the registry question is a market-structure question. A high threshold captures only hyperscale campuses; a low threshold captures co-location and mid-sized operators, redistributing compliance cost across the industry. The most contested filings in the docket will be about the threshold, because the threshold decides who is in the system at all.
The New-Category Governance Test
This is a Juncture working framework, first applied in this brief. It exists to assess regulatory orders that create a new actor category before its content exists.
Diagnostic question. What must be true for a newly created regulatory category to become a real governance instrument rather than a paper category?
| Element | Diagnostic question | Required evidence |
|---|---|---|
| Registry definition | Who is in the system? | Published registry criteria with thresholds stated, not deferred; a definition that captures the risk class without arbitrary size cutoffs |
| Obligation content | What must registered entities do? | Standards with specific, enforceable requirements, not process obligations |
| Single-point accountability | Is there a deadline that forces both instruments? | A fixed submission date with a named recipient that reviews, remands, or approves |
| Adjacent-track alignment | How does the category interact with neighboring regulatory tracks? | Documented alignment with tariff, interconnection, and planning instruments, or an explicit sequencing decision |
Decision sequence. Assess the registry definition first, because it determines the population. Then assess obligation content, because it determines what registration means. Then test the deadline and the review path, and finally check alignment with adjacent tracks. A failure at any earlier element cannot be repaired by a later one: a registry without standards captures entities for nothing, and standards without a registry bind nobody.
Failure modes.
- Threshold gaming: thresholds set so that the risk class can evade registration by being small, disaggregated, or served off the bulk-power system.
- Dual-track drift: the registry and the standards move on different timelines and the category is defined by whichever lags.
- Deadline slippage without consequence: an unmet deadline that produces an extension rather than a review, converting accountability into process.
- Paper category: a registry and standards that create obligations on paper and no compliance machinery, measured by reporting rather than behavior.
Cases where the test should not be applied. Orders that amend existing categories incrementally rather than creating new ones, where the population and obligations already exist; voluntary programs without registration or enforcement, where the test's accountability elements do not apply; and purely procedural delegations that create no new category.
Applied to the computational-load order as of August 2, 2026. The registry definition is not yet proposed. The obligation content is not yet drafted. The single-point accountability element is met: the December 31, 2026 deadline is confirmed in the FERC meeting summary opened this run. Adjacent-track alignment is unresolved. The test's answer at cutoff is that the category exists in law and has no content yet, which is the finding, not a gap.
Transparency caveat. This is a Tier 2 framework under development under Juncture's methodology. Based on this pattern, Juncture has developed a working framework to assess new regulatory categories. It explains recent outcomes but requires historical validation to demonstrate predictive power, and this brief is its first application, not its confirmation.
Resolvable outcomes
These are Juncture's assessment, not fact, and the three outcomes below are mutually exclusive and jointly exhaustive: exactly one will be true on the resolution date of April 30, 2027. Probabilities sum to 100 percent.
| Outcome by 2027-04-30 | Definition | Probability |
|---|---|---|
| On-time submission stands | NERC submits the Reliability Standards and registry criteria by December 31, 2026, and FERC has not remanded or rejected the submission by the resolution date. | 55% |
| Delayed submission | NERC does not submit the deliverables by December 31, 2026, or requests an extension that is granted. | 25% |
| Remand or rejection | NERC submits by December 31, 2026, and FERC remands or rejects the submission by the resolution date. | 20% |
On-time submission stands, 55 percent. A section 215(d)(5) directive with a fixed date is a forcing mechanism NERC has met before, the category is new but the drafting burden is bounded, and FERC review of a first submission typically runs beyond the resolution date, which leaves no remand on the record.
Delayed submission, 25 percent. The compressed calendar is real, the category has no precedent in the standards-development process, and stakeholder balloting on registry thresholds is likely to be contested, which is the classic cause of deadline slippage in NERC's process.
Remand or rejection, 20 percent. The submission must define a registry for a category with no established threshold practice, and if the standards lack enforceable content or the registry conflicts with existing registration categories, remand is the most likely FERC response.
Resolution rule for ambiguity. If the submission is late and FERC subsequently remands it, the outcome resolves to Delayed submission, as the earlier failure. If the submission is on time and FERC's review is still pending at the resolution date with no remand issued, the outcome resolves to On-time submission stands. Any FERC remand or rejection issued by the resolution date resolves to Remand or rejection regardless of the submission's timing.
Forecast record
| Field | Entry |
|---|---|
| Forecast timestamp | 2026-08-02 |
| Forecast horizon | 2027-04-30 |
| Resolution date | 2027-04-30, assessed within five working days of that date |
| Resolution authority | Juncture Policy editorial desk, on the public record only: NERC filings and work-plan publications, FERC orders and meeting summaries in Docket RD26-7-000, and the Federal Register. |
| Outcome definitions | On-time submission stands: NERC submits by December 31, 2026, with no FERC remand or rejection by the resolution date. Delayed submission: no submission by December 31, 2026, or a granted extension. Remand or rejection: on-time submission remanded or rejected by FERC by the resolution date. |
| Probability revision conditions | See the list below. |
| Update history | v1 2026-08-02, initial forecast. |
Probability revision conditions.
- Publication of NERC draft standards or registry criteria before December 31, 2026, revises according to content: clear thresholds and enforceable obligations revise toward On-time submission stands; contested or deferred thresholds revise toward Delayed submission or Remand or rejection.
- A NERC announcement of an extension request revises toward Delayed submission.
- Stakeholder filings in Docket RD26-7-000 showing broad opposition to proposed thresholds revise toward Delayed submission.
- Any FERC statement signaling a strict review posture revises toward Remand or rejection.
What to watch
- NERC's work-plan publications and draft standards during the second half of 2026, which will reveal the proposed registry thresholds.
- Stakeholder filings in Docket RD26-7-000, which will show which interests favor broad versus narrow registration.
- The December 31, 2026 submission deadline, and whether NERC meets it.
- FERC's review and any revisions after the NERC submission.
- Coordination with the six RTO and ISO large-load tariff responses under the June 18 orders, and the Department of Energy National Transmission Needs Study comment cycle.
Bottom line
For the NERC staff drafting the registry criteria, and for the data-center operator deciding whether to engage with the docket, the point is that the category is being defined in the next five months, and the definition will outlast everyone currently in the room. The December 31 deadline is real, and the registry threshold is the decision that matters most, because it decides who is in the system at all. Engage with the thresholds, watch the filings, and treat the submission deadline as the single point of accountability it was designed to be. The order created the category in law. The registry criteria will create it in fact.
Evidence and sources
Primary and institutional sources
- FERC, "Summaries | July 2026 Commission Meeting," July 16, 2026: ferc.gov. Opened this run. Source for the E-1 order in Docket RD26-7-000, the section 215(d)(5) directive, the Reliability Standards and Rules of Procedure deliverables, and the December 31, 2026 deadline.
- FERC, order E-1, Docket RD26-7-000: ferc.gov. [UNVERIFIED this run: order text referenced via the meeting summary; full order language not opened.]
- Department of Energy, "National Transmission Needs Study," July 9, 2026: energy.gov. Opened in heavy QA. Source for the September 7, 2026 public comment and consultation deadline.
- Juncture live tracker, "Large-load interconnection and tariff governance": juncturepolicy.org/ferc-large-load-governance/. The tariff and interconnection track; this brief is the reliability-governance institutional angle and is not a duplicate.
Load-bearing claim map
| Claim in this brief | Source |
|---|---|
| FERC ordered NERC under section 215(d)(5) to develop new or modified Reliability Standards for computational-load integration | FERC July 2026 meeting summary, opened this run |
| FERC directed NERC to revise its Rules of Procedure including registry criteria for computational-load entities | FERC July 2026 meeting summary |
| Both deliverables due December 31, 2026 | FERC July 2026 meeting summary (deadline confirmed in the opened summary text) |
| Order issued July 16, 2026 in Docket RD26-7-000 | FERC July 2026 meeting summary |
| June 18, 2026 large-load tariff orders to six RTOs and ISOs as the companion track | FERC July 2026 meeting summary; Juncture live tracker context |
| Registry thresholds and standards content | Not proposed at cutoff; labeled UNKNOWN and not asserted |
| Interaction with the tariff track | Unresolved; labeled as Juncture's institutional reading |
| DOE National Transmission Needs Study comment deadline of September 7, 2026 | Department of Energy, National Transmission Needs Study, opened in heavy QA |
| December 31, 2026 deadline as single point of accountability | Juncture assessment of the order's structure |
Publication cutoff: 2026-08-02. All sources last accessed 2026-08-02.