Ernst Responds to Markey’s SBIR Compromise, Narrowing the Distance — and Clarifying What’s Still Unsettled
WASHINGTON — January 14, 2026 — What has changed since our previous article focused on Senator Markey’s compromise draft is not the posture of negotiations — still serious, still incomplete — but the shape of the bargaining. On paper, it now looks less like one side talking past the other and more like two sides drafting in the same direction, then arguing about how far and how fast.
Dated January 13, Senator Joni Ernst of Iowa circulated a counterproposal in the form of an amendment in the nature of a substitute to H.R. 5100 (“To extend the SBIR and STTR programs, and for other purposes”). The substitute text itself is short‑titled the “SBIR and STTR Extension Act of 2026” (Draft Copy of MIR26019). These texts are available at the end of this article.
The difference in titles is not cosmetic. It signals two distinct strategies: Markey’s draft aims to be an expansive modernization of the statute, while Ernst’s latest offer is slimmer — more targeted, more time‑limited, and in several contested areas, notably more cautious. Yet slim does not mean static. In at least three areas central to the late‑2025 debate — foreign risk controls, major “strategic breakthrough” awards, and Phase III transition — Ernst’s newest text suggests movement toward a middle that would have been hard to see in earlier drafts.
The result is a negotiating moment that feels, in the best sense of the phrase, more legible. There is progress. There are also, plainly, a few hard questions left to settle.
Below we explore some of the core points of evolution in these negotiations.
The headline difference: permanence versus a three‑year bridge
The most consequential unresolved split is still the one applicants feel most viscerally: how much certainty Congress is willing to write into the program’s future.
Markey’s draft would remove the sunset mechanics by striking the SBIR termination subsection and deleting comparable STTR sunset language — a structural choice that, as written, would leave no new end date in those provisions. That approach matches the political argument many in the community have made for years: repeated cliff‑edge extensions are destabilizing, and the program should not need to be rescued on a rolling basis.
Ernst’s counterproposal goes the other direction: it extends SBIR authority by replacing “September 30, 2025” with “February 1, 2029,” and likewise changes STTR’s “fiscal year 2025” reference to “February 1, 2029” — a bridge of a little over three years. It also extends related authorities (“phase flexibility,” and other program activities) until that same date.
In practical terms, the gulf is not merely about philosophy. A permanent or open‑ended authorization changes how companies plan hiring, how investors assess risk, and how agencies build multi‑year pipelines. A three‑year extension, meanwhile, is easier to negotiate under time pressure — but it keeps the program on a countdown clock that would again demand congressional attention before long.
Still, even this gap contains a small note of convergence: both drafts are now working from the premise that Congress must do more than a patchwork continuation. The argument is over whether “more” means forever or for now.
Where the texts clearly converge: Phase III as the shared destination
If there is one theme that runs through both drafts — and reads like the emerging consensus in the room — it is that Phase III is no longer treated as a peripheral aspiration. It is treated as the point.
Ernst’s draft creates a dedicated “Phase III Award Education” section that would require workforce training for contracting officers and the acquisition workforce, with the text enumerating training topics including missions, goals, and authorities; use of Phase III agreements; Phase III data rights; and execution of Phase III sole‑source award contracts. The same section also contemplates funding for that training through amounts available under the commercialization readiness program subsection (y) or administrative/oversight funds under subsection (mm).
Markey’s bill contains a parallel Phase III education and training framework — similarly aimed at making Phase III less mysterious inside agencies and more routine in contracting culture.
Both drafts then push further, into what might be called process plumbing: the paperwork, templates, and institutional incentives that either speed Phase III adoption or quietly strangle it.
Ernst’s Section 6 would direct procurement center representatives to advocate “for the maximum practicable use and transition” of SBIR/STTR products into Phase III, and it would require SBA to update policy directives accordingly. It also adds a new obligation for agencies to develop simplified procedures and model contracts for Phase I, Phase II and Phase III awards, and to issue standardized solicitation provisions and contract clauses providing clear guidance on market‑research and eligibility information for Phase III.
Markey’s bill similarly adds Phase III award simplification requirements, including model contracts and standardized clauses, and includes a comparable emphasis on clarifying what companies can expect to provide for Phase III eligibility determinations.
In negotiation terms, this matters because Phase III is one of the few areas where neither side seems to view the other’s gains as their loss. Improving Phase III transition capacity is hard to spin as ideological, and easy to spin as practical.
The “strategic breakthrough” center of gravity — and its unsettled parameters
A second convergence is more ambitious: both drafts contemplate a small number of very large awards — the “strategic breakthrough” concept — intended to push select technologies further and faster than traditional Phase II levels.
But here, while the destination is shared, the route differs.
Markey’s approach: narrower allocation, explicit matching structure
Markey’s strategic breakthrough provisions cap the “strategic breakthrough allocation” at “not more than 0.25 percent” of the extramural budget for agencies above a threshold, beginning in fiscal year 2026. A “strategic breakthrough award,” in his text, could reach $30 million, last up to 48 months, and would be conditioned on a company demonstrating “not less than 100 percent matching funds,” with the further requirement that “not less than 20 percent” of that match come from government sources (not SBIR/STTR), and the remainder from outside private capital or other non‑SBIR sources.
In other words: limited carve‑out, high ceiling per award, and a structured match meant to ensure that “breakthrough” is tethered to real downstream demand — both public and private.
Ernst’s approach: larger ceiling on allocation, different match language
Ernst’s counterproposal defines a “strategic breakthrough allocation” as a required expenditure amount “of not more than 0.75 percent” of the extramural budget for eligible agencies, beginning in fiscal year 2026. Her draft likewise sets strategic breakthrough awards up to $30 million.
The matching structure, though, is written differently. Ernst would require a company to demonstrate “not less than 100 percent matching funds” from new private capital or “new funding awarded by a government agency under a program other than Phase I or II of the SBIR or STTR program.” The text does not impose Markey’s explicit “20 percent from government sources” floor; instead, it permits the match to be satisfied by either new private capital or new non‑SBIR government funding, as written.
For the Department of Defense, Ernst adds a particularly pointed gating mechanism: eligibility depends on a senior acquisition commitment — “a commitment for inclusion in a program objective memorandum from an official with the rank of program acquisition executive or higher.”
Why the difference matters
Both versions are clearly attempting to solve the same real‑world problem: the valley between “great Phase II prototype” and “adopted capability.” But they signal different instincts about risk.
Markey’s smaller allocation cap and required government portion in the match read like a safeguard against over‑concentration: the breakthrough lane exists, but with tighter guardrails and a built‑in test that at least one government buyer beyond SBIR is putting money on the table.
Ernst’s larger cap and match flexibility could, if adopted, create more room for agencies to shift SBIR dollars into a few big bets — potentially useful in a world where speed matters — while relying more heavily on agencies’ internal judgment and acquisition discipline (especially at DoD) to prevent misuse.
Negotiators may yet find that this is the kind of difference that yields to arithmetic: a compromise percentage, a hybrid match formula, or agency‑specific tailoring.
The foreign‑risk question: shared urgency, different due process
If Phase III is the shared destination and strategic breakthrough funding is the shared bet, then foreign risk is the shared anxiety — and the place where the two sides’ methods diverge most sharply.
Ernst’s counter: list‑based triggers and explicit “risk to national security” review
Ernst’s draft would require agencies to evaluate whether a small business concern presents a risk to national security “for any reason,” using tools including due diligence, disclosures, and coordination with intelligence and law enforcement. Her text then anchors “foreign risk” determinations to a series of named government lists — including the UFLPA Entity List, the Non‑SDN Chinese Military‑Industrial Complex Companies List, Section 889 list, and other enumerated lists in the text.
In the due diligence program itself, her counterproposal expands the “security risks” assessment to include items like cybersecurity practices, patent analysis, employee analysis, foreign ownership and obligations, research relationships (including co‑authorship), investment relationships, licensing agreements, joint ventures, and business relationships tied to a “foreign country of concern.” It would also require agencies to examine relationships to entities on specific lists “as published on the date of the closing of the solicitation.”
Notably, Ernst also includes a procedural element: when an agency denies an application on foreign‑risk grounds, it must provide a notification that (as appropriate and without compromising national security) advises the small business of the determination and identifies the criteria that formed the basis for it — and clarifies that a denial does not automatically bar eligibility in a later cycle.
Markey’s compromise: mitigation as the centerpiece
Markey’s due diligence approach keeps the core security screening structure but adds a different kind of friction — the kind meant to protect companies from being permanently defined by an addressable problem.
His draft adds a mitigation‑centered procedural step inside the due‑diligence program: when an agency determines that a small business concern poses a security risk, the agency must notify the firm of that determination before it formally notifies the firm of the award decision, and must provide the firm an opportunity to mitigate the risk.
Markey also extends the due diligence program authority through “September 30, 2045.”
The practical tension: speed and secrecy versus predictability and repair
Ernst’s model is comparatively straightforward for agencies: identify risk, rely on formal lists, document the determination, notify as appropriate. Markey’s model asks agencies to build — and staff — a mitigation pathway that may be more labor‑intensive, but also more legible to companies seeking to comply without being cast out.
Neither is inherently “soft” or “hard.” They are different answers to a common question: should SBIR security screening function more like an eligibility gate with limited feedback, or more like a compliance process that gives applicants a chance to correct course?
That is a negotiable difference, but it is not a trivial one. For companies, the presence or absence of a mitigation mechanism can change whether a security concern is experienced as a permanent brand or a solvable problem.
The “mills” debate shifts again: benchmarks, caps — and who bears the burden
Another place the two drafts circle the same issue from different angles is the long‑running concern about high‑volume award recipients.
Markey’s bill approaches the question through performance metrics. It defines a “covered small business concern” as one that has received more than 50 Phase I awards over the review period, and then requires such firms to certify—using the bill’s detailed “annual revenue” methodology—that a meaningful share of revenue is coming from Phase III awards or other non‑SBIR/STTR sources (25 percent over the review period, or 20 percent measured from the firm’s first SBIR/STTR award through its most recent fiscal year). If SBA determines those standards are not met, Markey’s text limits the number of Phase I applications the firm may submit in the next fiscal year to its total submissions over the review period divided by 3.5, with waivers available for mission‑critical or national security topics.
Ernst’s counterproposal takes a broader but potentially blunter tool: it requires each agency to “set a limit of a maximum number of proposals a small business concern may submit in a fiscal year,” for both SBIR and STTR solicitations.
The difference is not just philosophical. Markey’s mechanism is targeted — it applies to a defined group and is tied to commercialization outcomes. Ernst’s mechanism is universal — it applies across the applicant pool, with the numerical ceiling delegated to agencies.
A universal cap might reduce review burden quickly, but it also shifts risk onto early‑stage firms that rely on multiple proposals to find fit, especially across varied agencies and topics. A targeted cap, meanwhile, demands reliable data and careful implementation, and it can be contested on definitional grounds (what counts as revenue? how are affiliates treated?), even though Markey’s text supplies extensive definitions for “cash revenue” and related terms.
This is the kind of disagreement that often produces a compromise that looks technical: a baseline cap with exceptions, or a targeted cap paired with agency authority to impose broader limits under certain conditions. But as of now, it is still a fork in the road.
A quieter but meaningful convergence: technical assistance becomes more flexible
Technical and business assistance (TABA) rarely draws headlines, but it is often where policy becomes real for applicants — because it shapes whether a company can afford the help it needs to commercialize.
Here, the two drafts show notable overlap: both expand what TABA can cover, including cybersecurity assistance and screening for potential foreign involvement in development or commercialization activities.
But Ernst’s counterproposal makes a sharper structural change: it rewrites the statute’s opening language so agencies “shall authorize recipients” to select, if desired, technical and business assistance provided under the eligible uses — instead of emphasizing agency vendor agreements — and it strikes the existing “vendor selection” subparagraph and reframes the heading as “eligible uses of funds.” It then adds explicit authority to use assistance funding to hire or augment staff and to conduct training activities aligned with the program’s goals.
Markey’s approach also expands flexibility, but in a slightly different way: his bill keeps vendor‑based pathways and explicitly allows assistance to be provided through vendors selected under the statutory process, vendors other than those selected under that process, through staff activities, or combinations of those methods.
The direction is aligned — more flexibility, more modern categories of support — and the remaining question is how “open” the vendor ecosystem should be, and how agencies should oversee quality and conflicts without recreating the administrative burden everyone is trying to reduce.
Both bills also endorse Innovation Corps training (I‑Corps). Markey’s text applies to agencies that, as of January 1, 2026, run SBIR/STTR with I‑Corps and requires agencies to provide participation options and authorize use of funds for I‑Corps courses. Ernst’s language is similar in effect, requiring agencies with I‑Corps programs to provide the option and authorize use of subsection (q) funds for participation, and specifying permissible funding sources for participation costs.
What Ernst’s counter does not carry forward — and why that may be the point
To understand why Ernst’s January draft feels like progress, it helps to compare it with her earlier, more expansive bill: the original “INNOVATE Act” (S. 853, introduced March 5, 2025).
The INNOVATE Act was not merely an extension; it was an overhaul, and it contained several provisions that were widely expected to be difficult negotiating terrain.
Some of those provisions are now notably absent from Ernst’s counterproposal — and their absence is itself information.
The most consequential deletions: long lookbacks and recovery authority
INNOVATE defined “foreign risk” with an explicit 10‑year lookback—capturing conduct within the 10 years preceding submission of an SBIR/STTR application. Separately, it paired that posture with a recovery framework that could reach beyond 10 years where national security was deemed to require it.
Ernst’s counterproposal does not include those INNOVATE provisions. Instead, it emphasizes forward‑looking screening tools — list‑based triggers, expanded due diligence categories, and a notification process that (carefully) provides criteria without compromising national security.
That shift is a sign that the negotiation has moved from maximal enforcement posture toward a framework more likely to be administrable — and, therefore, more likely to be enacted.
The “volume” constraints become less prescriptive
INNOVATE would have capped proposal submissions with fixed numbers — including a limit of “3 proposals” per solicitation and “25 proposals” to an agency in a fiscal year.
Ernst’s counter replaces fixed statutory numbers with an agency‑set maximum, requiring each agency to establish a limit but leaving the ceiling itself to agency discretion.
This is not automatically friendlier to applicants — an agency could set a stringent cap — but it is meaningfully different from Congress hard‑coding a single nationwide number into statute. The negotiation space widens.
Eligibility and “new entrant” rules largely drop out
INNOVATE included a “funding cap” provision that would have barred a company from applying for additional Phase I or Phase II awards if it had received more than $75,000,000 in those awards. It also included other definitional and structural changes aimed at encouraging new entrants and limiting repeat participation, including a receipts threshold in the Phase I “small business concern” definition.
Those provisions are not part of Ernst’s January counterproposal. Instead, the counter’s approach to volume is the simpler proposal cap requirement and, in Markey’s draft, performance metric tools. The negotiation, in other words, is no longer centered on rewriting eligibility categories. It is centered on managing volume and risk within the existing structure.
Set‑aside increases: no longer in Ernst’s negotiating position
INNOVATE would have increased the SBIR set‑aside percentage (to 3.45 percent) and the STTR set‑aside (to 0.20 percent), as written.
Ernst’s counterproposal does not contain set‑aside increases. Markey’s bill does — dramatically — ramping the SBIR minimum for many agencies up to 7 percent over time and STTR up to 1 percent.
This remains one of the major substantive divides. And it is not surprising: raising set‑asides changes how agencies allocate billions in research dollars. It is harder to negotiate than process reforms, even when both sides agree on the value of SBIR/STTR outcomes.
The broader Markey agenda — and the narrower Ernst offer
One reason the two sides can now be described as “closer” is that Ernst’s offer is narrower, not because Markey’s offer has become smaller.
Markey’s bill still proposes a sweeping suite of additions: fellowships, rural outreach frameworks and definitions, expanded data collection, structural oversight changes, and technical modernization of the statute beyond security and Phase III transition.
For example, Markey creates SBIR/STTR fellowship authorities in Section 202. He defines “rural area” and adds outreach requirements aimed at rural communities. He also expands data collection requirements for awardees and research institution subcontractors. And he would prohibit political employees from participating in the selection of SBIR/STTR recipients — a “politicization prohibition” that signals a desire to keep awards insulated from political pressure.
Ernst’s counterproposal does not attempt to carry that entire modernization package. Instead, it concentrates on areas that have become the center of bipartisan agreement: security screening, Phase III transition, strategic breakthrough awards, and administrative burden.
This narrowing may be strategic. It may also be substantive. Either way, it changes the negotiation: it places greater weight on the question of whether a final compromise will look more like a comprehensive reauthorization bill (Markey’s model) or a targeted extension with select reforms (Ernst’s model).
So where does this leave applicants?
If you are an applicant looking for the simplest possible update, it is this:
A week ago, the debate still felt like it was stuck in competing universes of what reauthorization should be. Now it feels like it is in competing versions of the same bill.
Both sides are writing toward:
- stronger and more explicit foreign‑risk screening mechanisms (with different approaches to mitigation and communication),
- an expanded strategic breakthrough pathway for unusually large awards (with different caps and match structures),
- and a Phase III push that is no longer rhetorical but structural — education, policy directives, model contracts, standardized clauses.
The remaining gaps — permanence vs. a 2029 bridge, broad funding expansions vs. a narrower approach, and targeted benchmarks vs. universal proposal caps — are serious, but they are also the kind of gaps that can be bridged by negotiated numbers, sunsets, waivers, and carefully drafted agency discretion.
That is not a guarantee of a deal. It is, however, what progress looks like in a negotiation that has finally begun to exchange text rather than talking points.
And for a community that has spent months watching the calendar more than the substance, the existence of competing text — real text, in legislative counsel format, with sections that can be compared line‑by‑line — is itself a sign that the process is moving again.