Expert advice for every grant — from EPA to DARPA

On May 19, the Department of Energy's Office of Critical Minerals and Energy Innovation announced 19 selections under FOA 3105 — two pilot-scale facilities for magnesium and rare-earth separation, and 17 technology development projects spanning lithium, cobalt, nickel, graphite, silicon, and manganese. Here is what the selection list tells researchers and small businesses about where DOE wants the supply chain in 18 months — and where the next solicitations will go.

The Legal Services Corporation's Technology Initiative Grant cycle for calendar-year 2026 closed pre-applications on April 10 and opened a new $75K Planning Grant category. Full applications for the General TIG and SEA categories are due June 30. The 2024 award list — 32 grants, $5M+, dominated by AI chatbots, document automation, and Copilot deployments — is the clearest signal of what LSC is buying with TIG money and how legal-aid organizations should position their 2026 submissions.

NSF 26-508 funds one State/Territory AI Coordination Hub per jurisdiction at $1M per year for three years — up to 56 awards and $224M total. Only one proposal per institution. Round 1 LOIs are due June 16, 2026 and full proposals July 16. The structure will determine whose convening capacity defines AI workforce strategy in every U.S. state for the rest of the decade.

On May 11, SCO Director Jay Dryer publicly described the office's three-portfolio structure (long-range fires; autonomy and AI; special and enabling capabilities) and eight focus areas spanning precision fires, contested logistics, kill webs, and cost-effective air defense. SCO is executing a $1.7 billion FY 2026 budget through a standing Broad Agency Announcement that accepts white papers through August 31, 2029. The office is not DARPA, the engagement model is different, and the winners are not who you would expect.

S. 3971 — the Small Business Innovation and Economic Security Act — reauthorized SBIR and STTR through September 30, 2031 after a six-month lapse. The legislation adds Strategic Breakthrough Awards up to $30M with 100% matching, eight-watchlist foreign-affiliation screening, and FY 2027 per-company proposal caps. Companies that built their pipeline around volume submissions need a new strategy now.

The National Institute on Aging's FY 2026 AD/ADRD portfolio consolidates the dementia research infrastructure layer — NCRAD, NACC, the new AI and Technology Collaboratory Coordinating Center — into a small number of large, often single-source cooperative agreements. The $113M new-research increment goes elsewhere. For investigators submitting in FY 2026, the structural change matters more than the headline dollar number.

NSF 26-503 replaces the long-running CyberCorps Scholarship for Service with CyberAICorps — a dual-authorized program written against two statutes that explicitly fuses AI competency into the federal cybersecurity workforce pipeline. The July 21, 2026 deadline is the first chance to compete under the new framework, and the $2.5M Scholarship Track and $500K Innovation Track each have constraints that will determine which institutions get a foothold.

On May 8, 2026, ED and HHS announced the first competitions under a new Family Engagement and School Support Partnership — covering Promise Neighborhoods and Ready to Learn, with applications due August 6. HHS will manage grant funds and provide technical assistance under an Interagency Agreement, the largest structural change to Promise Neighborhoods since its 2010 launch. The priority shift to evidence-based literacy and high-impact tutoring is the visible piece. The administrative reorganization is what determines whether the program survives the next reauthorization.

ARPA-H launched the Intelligent Generator of Research program on May 5, 2026 with only three anticipated OT awards, a Solution Summary deadline of June 25, and a full proposal due August 6. The program is also piloting LLM-assisted proposal review. Here is the strategic read.

DARPA-PS-26-04, published February 25, 2026 by the Tactical Technology Office, restructures the contract around three phases — Phase 0 Backbone (6 months), Phase 1 Base (12 months), Phase 2 Option (18 months) — and culminates in an instrumented flight-test campaign. The solicitation is not really about T&E. It is about the digital-twin and uncertainty-quantification middleware DoD needs for any AI-enabled combat system.

DARPA's Mathematics of Boosting Agentic Communication program — DSO-led, $2M Phase I cap, abstracts already in, full proposals due June 16, 2026 — is the first federal initiative to treat multi-agent AI communication as a mathematical object rather than a product feature. The Mendeleev-rediscovery benchmark in the solicitation is the tell.

Ten foundations co-funded $18M to shape AI for the public good — $8M to twelve inaugural grantees at $500K each, $3M to AI Civics, and $10M reserved for a summer 2026 open call. The grantee list and the absences are both strategic signals.

NIH consolidated more than 300 Notices of Funding Opportunity off the books in 2025 and is pushing investigators toward parent announcements in FY2026. The framing is administrative simplification. The mechanism is a structural shift in which research areas get advertised and which do not — and which investigators get rewarded for ignoring NOFO topics entirely.

The Commerce Department's August 2025 march-in proceeding against Harvard is the first invocation of an authority that sat dormant for 45 years. The policy precedent reaches every Bayh-Dole grantee — and the operational compliance gap is wider than most institutions realize.

The DOE Genesis Mission RFA closed its Phase II window on May 19. With $293.76M, 21 topics, and 99 focus areas, it is the largest single federal AI-for-science procurement in 2026. Here is what survived the cut and what comes next.

NASA's SBIR/STTR Program Year 2026 abandons the annual solicitation in favor of a rolling BAA. Phase I awards jump to $225K and Phase II to $1.275M. Here is the playbook.

NICHD's FY2026 funding strategy applies an automatic 14 percent reduction to every new R01 below the peer-review recommended level, eliminates inflationary increases on future-year commitments, and abandons a fixed payline entirely in favor of priority-driven discretion. The structural implications for child health investigators.

William Penn's 128-grant, $57.2M May 2026 distribution reveals a Philadelphia-focused funder doubling down on children, arts education, and civic infrastructure as federal support recedes.

NSF's new K-12 innovation foundry closes its planning round May 27, 2026, with up to 50 awards of $50K each. The mandatory four-role team — educator, technologist, researcher, parent — is the binding constraint.

OPM's March 2026 rule strips civil service protections from positions deemed policy-influencing — including potentially thousands of grant reviewers and program officers across NSF, NIH, and the Department of Defense. Here's how grant seekers should adapt.

DOL and ED announced May 21 a one-time, $366 million Title III SIP competition — more than triple the $102 million Congress appropriated — by folding reallocated Minority-Serving Institutions and Hispanic-Serving Institutions funds into a single pool. Here is what every eligible college needs to know about the three competitive preferences, the workforce Pell connection, and how to position by June 23.

The May 19, 2026 final rule extends Pell Grants to 8-15 week workforce programs — but only those certified by governors as high-skill, high-wage, or in-demand, with at least 70% completion and 70% job placement. Here is what eligible colleges, workforce boards, and non-college providers should do before the July 1 launch.

ARPA-H's Intelligent Generator of Research program targets the gold-standard replication problem with a 5-year, 3-phase initiative awarding roughly three Other Transaction agreements. Solution Summaries due June 25, 2026. Full proposals August 6. How interdisciplinary teams should position now.

Humanity AI's first $18M in pooled grants — $500K to 12 organizations plus $3M to Data & Society — signals where the ten-foundation coalition's $500M will land over five years. Here's the strategy behind the picks and the open-call window.

Judge Colleen McMahon ruled on May 7 that DOGE's mass termination of 1,400 NEH grants violated the First and Fifth Amendments. The order rescinds termination letters but does not force payment. What humanities organizations should actually do in the next 90 days.

NIH's NOT-OD-25-132 bars applications 'substantially developed by AI' — but ten months in, the working rule is disclosure, not a tool ban. Here is what counts.

NIH's Malign Foreign Talent Recruitment Program certification leniency window closes in May 2026. After it ends, every senior/key person on every NIH RPPR must clear SciENcv validation or block the report. The compliance details PIs are missing.

NSF's new X-Labs initiative — $1.5B over a decade, $50M/year Phase 1 awards, OTA mechanism, July 13 deadline — is a structural break from the standard cooperative agreement model. Here's what it means for teams chasing the first two topics.

On May 12, USDA Rural Development adopted the department's modernized NEPA framework at 7 CFR Part 1b, retiring 7 CFR 1970 and changing how environmental review interacts with every RD grant application. The technical changes, the transition dates, and how applicants should adjust their timelines.

The Department of Education's redesigned Comprehensive Centers Program competition opened May 8, 2026 with $46.015M for roughly 19 awards. Three tracks — National, Regional, Content — and a redesigned mandate to align with state priorities. Here is how the field has shifted.

DARPA BTO pre-released four FY26 SBIR/STTR topics on April 30, 2026, with proposals due June 3. Two topics — SWiFT and EXPOSITION — offer Direct-to-Phase-II awards up to $1.5M, bypassing the standard Phase I gate. Here is what each topic is actually solving, why the DP2 structure matters, and how small biotech, surgical robotics, and battlefield-medicine teams should decide whether to compete.

EPA's Gulf of America Division announced up to $50 million for the Farmer-to-Farmer grant program on May 5, 2026, with 20–30 awards of $1.5M to $2.5M each across EPA Regions 3–8 and a June 19, 2026 deadline. The funding rewards farmer-led organizations that can demonstrate working-lands conservation at scale. Here is how the eligibility, partnership structure, and watershed geography actually decide the awards.

The joint NSF-NIH Smart Health and Biomedical Research solicitation supports high-risk, high-reward AI/data science work in health — $300K per year for four years, with 20+ NIH institutes participating. Here is how the program actually selects winners.

On April 13, 2026, President signed S. 3971 reauthorizing SBIR/STTR through September 30, 2031 — the longest extension in program history. The new law introduces $30M Strategic Breakthrough Awards, FY2027 proposal caps to address SBIR mills, expanded foreign-risk screening, and stronger TABA support. Here is how small businesses, university spinouts, and dual-use startups should reposition for the next two solicitation cycles.

The Workforce Pell final rule hit the Federal Register on May 19, 2026. Programs of 150-599 clock hours, 8-15 weeks, with 70% completion and placement thresholds qualify — but only with Governor and Department approval. Here is what institutions must do now.

ARPA-H disclosed in the IGoR solicitation (ARPA-H-SOL-26-155) that secure LLMs will assist initial review of submitted materials — a direct departure from NIH's ban on generative AI in peer review and NSF's pending working-group deliberations. The policy split signals where federal proposal review is headed and what applicants need to do differently.