NIH Just Killed the K-Award Data Sharing Plan a Year Early — and the Notice Covers Applications Already Sitting in Review

Most NIH burden-reduction notices arrive with a future effective date and a transition table. Applicants read them, file them, and come back in a year. NOT-OD-26-120, released October 5, 2026, did the opposite: it took a change already announced for October 12, 2027 and made it effective immediately, retroactively covering applications that are already submitted and awards that are already active.

The change itself is narrow. Data Management and Sharing Plans are no longer required for NIH Career Development (K) awards. The implementation mechanics are what make it unusual — and what make it worth reading carefully if you have a K application in the queue right now.

What the Original Announcement Said

On August 12, 2026, NIH issued NOT-OD-26-098, a wide-ranging overhaul of the Career Development award portfolio. The headline items were structural: collapsing more than a dozen individual K activity codes into a streamlined set of parent notices of funding opportunity organized by program type, clarifying allowable uses of K funds for clinical trial activities, and — the item that drew the most attention — eliminating the independent clinical trial K.

Buried in that notice was a smaller provision: K awards would no longer require a DMS Plan, for application due dates on or after October 12, 2027.

NIH's stated reasoning was clean. K awards fund career development. They do not provide primary support for a research project. Requiring a full data management and sharing plan from a mechanism whose output is a trained investigator rather than a dataset was an alignment problem, not a policy problem.

But October 12, 2027 is a long way off. Every K applicant between August 2026 and that date would still have been writing plans, and every active K recipient would still have been updating them in progress reports, for a requirement NIH had already conceded did not belong there. We covered the broader restructuring in our analysis of the K award consolidation into KR1, KC1, KS1, and KT1, where the October 2027 date governs nearly everything else.

NOT-OD-26-120 pulls this one piece out of that timeline and lands it now.

The Three Cohorts, Spelled Out

The notice does something NIH policy documents frequently do not: it names every population and tells each one explicitly what it no longer owes.

Active K award recipients are not required to submit or update a DMS Plan, including through the Research Performance Progress Report.

Applicants with pending K applications — meaning already submitted and sitting in review or pre-award — are not required to submit a DMS Plan or to provide an updated or revised plan as part of the pre-award process, including at Just-in-Time.

Applicants submitting future K applications are not required to submit a DMS Plan at all.

Then the sentence that does the actual legal work: NIH will not consider the absence of a DMS Plan to be noncompliant with the DMS policy.

That line matters more than it looks. The 2020 Final NIH Policy for Data Management and Sharing (NOT-OD-21-013, effective for applications submitted on or after January 25, 2023) made plan submission a term and condition of award. Removing a submission requirement is not the same as removing the compliance exposure that attaches to a missing document. NOT-OD-26-120 closes that gap in one sentence rather than leaving institutions to infer it.

The Just-in-Time carve-out is the single most immediately useful provision. JIT is where pending applications get their DMS Plans scrutinized and revised, often under a short turnaround from a program officer. Anyone in that window as of early October 2026 just had a task deleted mid-flight.

The Collision With July's Notice

Here is where K applicants and their research administrators are most likely to reach the wrong conclusion.

On July 29, 2026, NIH issued NOT-OD-26-100, which eliminated the prior-approval requirement for modifying an approved DMS Plan and relocated change reporting into RPPR Section C.5.c. That notice took effect October 1, 2026 — four days before NOT-OD-26-120 — and it was explicit that training and fellowship awards were carved out while Career Development K awardees were required to comply. We walked through that trade in detail in our piece on the retirement of DMS prior approval.

So within a single week, NIH published guidance affirmatively naming K awardees as subject to DMS reporting, and then guidance stating K awardees have no DMS Plan to report on. Both notices are live. Neither has been withdrawn. The later one governs, and NOT-OD-26-120 says so by implication — but only if you know to read them in sequence.

An administrator who pulls up NOT-OD-26-100 in November, sees "K awardees are required to comply," and sends a chase email to a K recipient about an empty Section C.5.c will be enforcing a requirement that no longer exists. That is not a hypothetical failure mode. It is the predictable result of two burden-reduction notices landing four days apart on overlapping ground.

What Did Not Change

Four things survive intact, and conflating any of them with the DMS removal is the expensive mistake.

Everything else in NOT-OD-26-098 holds its original dates. The notice is explicit: all other applicable provisions and dates remain unchanged. The K consolidation, the parent NOFO restructuring, and the elimination of the independent clinical trial K still key to October 12, 2027. Only the DMS provision accelerated.

Linked research awards keep their own obligations. A K99/R00 transition, a mentored K with an associated R-mechanism parent award, or a KL2 scholar working under an institutional CTSA grant — the underlying research award carries its own DMS Plan and its own reporting. NOT-OD-26-120 removed a requirement from the K, not from every award a K scholar touches.

Other data-sharing authorities are untouched. The NIH Genomic Data Sharing Policy, model organism and resource sharing expectations, clinical trial registration and results reporting under FDAAA and NOT-OD-16-149, and repository-specific deposition terms all operate on separate authority. A K scholar generating human genomic data still has institutional certification and registration obligations that have nothing to do with the DMS Plan.

NIH's stated commitment to data sharing is unchanged. The notice says so directly. Removing a plan requirement from a career development mechanism is a scoping decision, not a retreat. Reviewers can and will still read a candidate's approach to data as part of the research training plan — it simply is not a separately attached, separately enforceable document.

Where the Paperwork Will Lag

The notice closes with standard language: NIH will update applicable funding opportunities, application instructions, award guidance, and other relevant materials as appropriate.

"As appropriate" is doing a lot of work there. Between a policy notice and a cleaned-up form package, there is typically a gap measured in months. For K applicants with fall and winter due dates, that gap creates three practical questions:

Does the NOFO you are applying under still list a DMS Plan as required? Almost certainly yes, for now. Parent K announcements and institute-specific K NOFOs were written before October 5. The notice supersedes them on this point, but the text in front of you will not say so.

Will ASSIST or Workspace throw a validation error on a missing attachment? If the DMS Plan field is configured as required in the current form package, it may. The resolution is a call to the eRA Service Desk or your assigned program officer with NOT-OD-26-120 cited by number — not a decision to write a plan you no longer owe.

For pending applications, should you withdraw a plan already submitted? No. The notice says you are not required to submit or to provide an updated or revised plan. It does not instruct anyone to remove one. An already-attached plan is harmless; the relief is that nobody can make you revise it at JIT.

The low-cost move for anyone with a K in flight this fall is to keep the notice number in the application file and in the JIT correspondence folder. When a reviewer, a sponsored programs officer, or an institutional compliance checklist asks for the plan, the answer is a citation, not an argument.

The Direction of Travel

Read alongside the rest of 2026, NOT-OD-26-120 fits a consistent pattern. NIH has spent the year stripping transactional steps out of its application and reporting machinery: letters of support replaced by standardized letters of collaboration under NOT-OD-26-094, the DMS Plan format itself cut down to a short structured instrument, prior approval gates retired, and now an entire document class removed from an entire award family.

The difference with this one is the direction of the burden transfer. In most of the 2026 notices, NIH removed a step it used to perform and relocated the corresponding recordkeeping to the recipient. NOT-OD-26-120 does not do that. It deletes the obligation on both sides. For K candidates — early-career investigators with the least administrative support and the most to lose from a JIT delay — that is an unambiguous gain, and it arrived twelve months earlier than anyone writing a K application in September had any reason to expect.

The applicants who will not capture the benefit are the ones who never see the notice, write the plan anyway because the NOFO still asks for it, and spend a week in October on a document NIH has stopped reading.

Granted tracks NIH policy notices alongside the funding opportunities they govern, so a change in what your application must contain surfaces in the same place as the deadline for filing it.