DOE DE-FOA-0003612: The Genesis Mission: Transforming Science and Energy with AI (FY26–FY27)

DOE DE-FOA-0003612: The Genesis Mission: Transforming Science and Energy with AI (FY26–FY27)

At-a-glance snapshot

What this opportunity is and who it is for

DE-FOA-0003612 is a federal RFA from DOE for teams that can combine AI with experimental science, national lab resources, and mission-relevant industry or academic expertise. The goal is to push scientific workflows and R&D cycles faster in selected challenge areas spanning advanced manufacturing, biotechnology, critical materials, nuclear, quantum, semiconductors, discovery science, and energy.

The RFA is unusual in one key way: it is explicitly designed for mixed collaboration formats. The language around lead institutions, partner institutions, and required teaming is not a generic grant formality. It is part of a larger mechanism to force a practical division of labor between institutions that generate science and institutions that can scale, test, and industrialize outputs.

For applicants, this matters because the “competition” is not just about a good idea. It is a systems test of execution architecture. The FOA language places strong emphasis on team composition, data requirements, and project readiness. Teams that treat this as a simple one-lab proposal are typically rejected at the review or compliance level before scientific merit is judged.

Because the solicitation explicitly remains connected to Phase I and later Phase II tracks, it is especially useful for teams that can run in two stages: first prove viability in a concentrated nine-month Phase I, then apply for a larger and longer execution team in Phase II. That structure is why the same FOA page is relevant beyond the first two 2026 deadlines, and why many teams monitor it through late 2026.

Why this FOA is currently relevant for 2026–2027 planning

There are three practical signals that make this one time-sensitive and worth adding now:

1. The FOA itself was posted in March 2026 and later amended in April with corrected submission timing and updated collaboration language.
2. The main close date for FY26-linked funding activity is in December 2026, while internal stage deadlines earlier in 2026 still govern Phase I and first-wave Phase II entries.
3. The solicitation explicitly notes continuation into later phases and indicates DOE may issue follow-on structures for FY27.

This is not a legacy page for historical reference only. Even though the official notice is tied to FY26 in its current form, the sequencing and pipeline language indicates continuity toward FY27 opportunities. Teams that miss the first-stage application windows can still align to the continuation path if they already understand the requirements, because Phase I projects can be advanced toward Phase II within the same solicitation framework.

The most useful distinction is this: this is not one-off micro-funding for a narrowly defined grant category. It is a platform-like RFA for AI-accelerated mission science. That matters because teams should plan around institutional roles, data strategy, and multi-year execution readiness, not just a one-off grant write-up.

Funding details you can rely on

The FOA gives a broad total budget signal: about $293.76 million in DOE support for this solicitation framework.

For initial planning, you can use the following working model:

• Phase I awards are defined as smaller, high-velocity projects with a typical size range of $500,000 to $750,000.
• Phase I projects have an expected performance period of about nine months.
• Phase II is intended as a larger follow-on path, expected at roughly 3 to 5 times the Phase I funding per year.
• The overall Phase II window is much longer, with three-year periods in active planning.

When teams evaluate fit, a common mistake is to anchor their proposal on an over-ambitious 3-year execution plan already in Phase I. A cleaner approach is to set realistic Phase I milestones that can be reviewed quickly, then justify the need for Phase II scale in the continuation narrative. Reviewers and management teams generally evaluate whether the architecture, team quality, and AI-method workflow can withstand scale-up.

The total pool is large, but award count is demand-driven. The FOA is clear that the exact number of Phase I and Phase II awards is dependent on both appropriated funds and project merit. That means competitiveness is real: even strong projects must be procedurally complete and review-ready.

Eligibility and teaming rules: what can disqualify you quickly

Before writing anything, map your team to the mandatory teaming format.

Required team nature

The FOA states that multi-institutional teams are required. That means a single organization submission model is not the right starting point unless your internal structure effectively includes multiple institutions through formal partner entities.

For FY26 Phase I, teams must include institutions from at least two of these three categories:

• DOE/NNSA national laboratory partner(s)
• Industry partner(s)
• IHE/non-profit/other partner(s)

For FY26 Phase II, the minimum shifts toward a stronger DOE-linked and industrial base:

• At least one DOE/NNSA national lab
• At least one industry partner
• IHE/non-profit/other partners are strongly encouraged

The practical implication is clear: if your submission is primarily university-based research with no industrial partner, the package is not just weaker, it may miss the structural requirement in explicit FOA sections.

Consortium membership

The FOA includes Genesis Mission Consortium members as eligible participants, but consortium membership is explicitly not required to apply. That is important. You should not delay submission while waiting for specific consortium invitation criteria. Build on your existing institutional network and use consortium channels for outreach and information, not as a prerequisite.

Domestic and subrecipient logic

The FOA allows proposing non-domestic subrecipients in applications, but this comes with strict justification expectations. Teams should be explicit about why international partners add unique capability and why equivalent domestic capability is not available. If you list foreign collaborators, reviewers and compliance staff will likely test whether this is essential rather than optional.

Limits on late and informal submissions

The FOA allows “additional applications” in some circumstances after deadlines but explicitly says DOE may decline them without review. For planning, treat this as a formal “do not rely on extension logic” rule.

Submission structure and deadlines you must plan around

The FOA has multiple deadlines and stages, and confusion here is one of the highest rates of avoidable failure.

Core official deadlines

• FY26 Phase I applications: May 1, 2026 (11:59 PM ET)
• FY26 Phase II letters of intent: May 1, 2026 (5:00 PM ET)
• FY26 Phase II applications: May 19, 2026 (11:59 PM ET)
• FY26 Phase I-to-Phase II continuation applications: December 17, 2026 (11:59 PM ET)

The page shows these as the operative schedule for the active FOA version. In practice, teams should still watch for further amendments and additional notices from DOE until final close.

FOA amendment context

The version currently listed is Amendment 000003. In practical terms, this is important because earlier dates and procedures were modified and must not be used as standalone truth. Always reference the current amendment page and the official PDF for the latest process details before submission.

LOI and direct phase entry logic

For FY26, DOE uses LOIs for many Phase II workflows. The FOA says LOIs are strongly encouraged for Phase II, and a missing LOI can affect how quickly DOE responds. Teams targeting Phase II should treat the LOI as an operational step, not a courtesy form.

Some teams may apply directly to Phase II rather than passing through Phase I, but this is only appropriate for teams with strong maturity and direct relevance.

Important operational rule: submissions channels

The FOA is explicit about channel requirements:

• Use the Grants.gov path tied to the RFA/assistance listing.
• Do not use FedConnect for FOA applications.
• Do not submit proposals through PAMS.

This is a hard compliance boundary. If your grant software stack is oriented around an incorrect portal, you risk technical rejection before evaluation.

Application system, preparation sequence, and pre-submission readiness

If you have no UEI and no complete SAM.gov registration, start there first. DOE states that registration can take weeks. Practical teams should begin that process well before the May window or before they begin assembling institutional letters and budget attachments.

A robust sequence that aligns with the FOA structure is:

1. Registering legal entity in SAM and generating a valid UEI.
2. Confirming team letters and roles (lead institution + partners), including PI and administrative points.
3. Confirming submission system access in Grants.gov using the correct assistance listing.
4. Creating a proposal skeleton that separates mandatory content from optional attachments.
5. Preparing a concise key-scope narrative around team composition and challenge fit.
6. Running a compliance pass against the required content list and deadlines.

What to prepare before the proposal write period

Even before you have all technical content, teams should prebuild four blocks:

• Team map: lead institution, PI, co-PIs, partner institutions, and role language.
• Challenge-to-method mapping: identify your precise AI intervention in the scientific workflow.
• Budget architecture: phase-appropriate spending model (Phase I versus Phase II expansion logic).
• Evidence package: data sources, infrastructure access plan, and partner access commitments.

The FOA is explicit about expected clarity on project management readiness, budget periods, and partnerships. Teams that leave these as afterthoughts often fail on responsiveness.

What reviewers usually want to see

While the full review criteria are in the solicitation, the practical expectations can be inferred from how this FOA is written.

Successful applications generally demonstrate:

1. A concrete science bottleneck and a clear AI intervention.
2. A team architecture that is genuinely cross-institutional.
3. A plan to deliver in nine months for Phase I while preparing for scale-up.
4. A budget where direct costs and effort are tightly tied to milestones.
5. A risk framework that explains why this approach is not merely computationally interesting but operationally impactful.

In this solicitation, review quality is inseparable from execution architecture. A high-quality algorithm idea without a credible data stewardship plan and partner workflow is weaker than a lower-profile method with strong deliverability.

Practical proposal writing strategy

Teams generally improve acceptance odds by splitting narrative strategy into three writing tracks.

Track 1: challenge and impact framing

Use plain language tied to a named focus area and the mission target. Explain what scientific or operational step is currently bottlenecked and how AI changes speed, precision, or discovery throughput. Avoid broad claims like “transformational research” unless you can show a specific technical chain.

Track 2: team and execution plan

List who does what:

• Lead institution governance and budget ownership.
• DOE/NNSA lab coordination with access to infrastructure, data, and computation.
• Industry contribution (prototype integration, deployment path, or commercialization context).
• Academic or non-profit contribution (algorithm development, test methods, workforce, unique datasets).

The execution plan should show that these are not names on paper. Include ownership and sequence.

Track 3: timeline and output package

For Phase I, use a 9-month logic. Make sure the output set is realistic: what is delivered by each month, what compute needs are requested, what external dependencies exist, and what the go/no-go path looks like for larger follow-on.

For teams applying toward Phase II continuation, connect the Phase I output directly to a Phase II architecture. If your Phase I outputs are not scale-ready, reviewers may not support a larger award.

Common mistakes to avoid

Treating this as a single-investigator grant

The solicitation is built for teams. If your “team” is only nominal and not institutionally complete, it is likely to score poorly on responsiveness.

Missing mandatory channel and registration steps

Submitting through a non-GrantGov route or with incomplete SAM/UEI status is a preventable failure.

Confusing deadlines across phases

May 1, May 19, and December 17 are different submission logics and different applicant intents. Submitting the wrong package against the wrong deadline is a frequent avoidable error.

Weak partnership justification for non-domestic subrecipients

The FOA allows non-domestic entities in some cases, but only with clear capability justification. If this justification is thin, reviewers and compliance staff will treat it as a risk flag.

Underestimating post-award complexity

Phase II work is often more than “more money.” It requires stronger governance, reporting cadence, and deliverable control. Teams should avoid presenting Phase II as a direct extension of Phase I without additional planning detail.

Ignoring post-amendment text

Because amendments already changed key dates and teaming framing, teams must use the latest notice text, not older drafts copied by hand.

Frequently asked practical questions

Is this only a DOE 2026 opportunity?

The solicitation title and issue year are FY26, but the document text and close structure indicate a pipeline that continues into FY27 through continuation pathways. It should be treated as active planning input for both 2026 and 2027 cycles.

Can applicants apply directly to Phase II?

Yes, the FOA supports direct Phase II applications in the defined pathways. However, this is a stronger governance model and should only be attempted by mature teams with a solid readiness case.

Do teams need a LOI for Phase II?

The FOA says LOIs are strongly encouraged for Phase II and sets a deadline. Skipping the LOI should be treated as a delay risk, not as routine strategy.

Is consortium membership required?

No. Membership can be helpful but is not a submission prerequisite.

Where is the best official source?

Use the DOE FOA detail page and the linked official PDF for amendment text, review requirements, and application instructions.

180-day planning checklist before submission

If your team is serious about this FOA, use this practical checklist:

1. Confirm team composition against Phase I and Phase II minima.
2. Align the RFA’s challenge area and expected deliverables with your AI intervention.
3. Finalize institutional roles, legal lead, and budget authority structure.
4. Complete SAM/UEI and verify applicant profile in Grants.gov.
5. Draft application templates and title page fields exactly.
6. Build a response pack for reviewers: method, data access, execution risks, scalability plan.
7. Run a compliance sweep for required submission documents.
8. Back-map timeline to the exact Phase I/Phase II deadline sequence.
9. Submit with enough buffer to catch portal errors.

The final practical signal for this opportunity is not whether your idea is compelling in isolation, but whether your team can turn it into a compliant, review-ready, multi-institution execution package.

Official sources and next steps

Use the official documents directly as your primary source before final draft:

• FOA landing page (official list and dates): https://science.osti.gov/grants/FOAs/FOAs/2026/DE-FOA-0003612
• Official PDF amendment package: https://science.osti.gov/-/media/grants/pdf/foas/2026/DE-FOA-0003612-000003.pdf
• Registration and submission channels referenced on the FOA landing and PDF (Grants.gov, SAM.gov)
• DOE announcement context: https://www.energy.gov/articles/energy-department-announces-293-million-funding-support-genesis-mission-national-science

Bottom line

DE-FOA-0003612 is most useful for teams that have both scientific depth and collaboration capacity. The opportunity is designed for interdisciplinary teams that can show strong AI-enabled scientific progress and then scale that work through a larger follow-on structure. If your team is only one institution with weak partner roles, start by building team architecture before drafting. If your team is already aligned, use the FOA’s staged structure: complete SAM/UEI, submit cleanly through Grants.gov, and treat the LOI/application path as a formal process, not a template exercise.