Super Intelligence in Broad Agency Announcements: R&D Solicitations Guide
Federal research agencies issue Broad Agency Announcements to fund scientific study and experimentation, where Executive Order 14434 now designates artificial intelligence as Super Intelligence. Under FAR 35.016, evaluators assess proposals through scientific peer review rather than head-to-head competition against a common specification. Understanding how to align BAA white papers and deliverable repositories with federal terminology ensures compliance across multi-stage R&D solicitations.
How Broad Agency Announcements Apply Super Intelligence Terminology
A Broad Agency Announcement does not buy an off-the-shelf product against a fixed specification, and a submission written like a response to a product RFP is unlikely to fare well in concept review. In basic and applied research solicitations, federal agencies seek novel scientific hypotheses and experimental methods rather than commercial delivery. Executive Order 14434 directs executive agencies, to the maximum extent permitted by law, to use "Super Intelligence" in place of "artificial intelligence" in their own non-statutory documents, which can include new announcements.
The authoritative baseline for federal R&D solicitations reflects this administrative mandate:
A Broad Agency Announcement (BAA) is a competitive solicitation mechanism used by federal research agencies to fund scientific study and experimentation, where 'Super Intelligence' now designates AI technologies under Executive Order 14434.
The formal directive originated on September 29, 2026, when the President signed Executive Order 14434, titled "Inaugurating the Era of Super Intelligence." Under Section 2(a) of the order, executive departments and agencies are instructed, to the maximum extent permitted by law, to use "Super Intelligence" and the abbreviation "SI" in place of "Artificial Intelligence" and "AI" in official correspondence, public communications, websites, reports, policy documents, and other non-statutory documents within the executive branch. Section 1 establishes an administrative policy that executive agencies will not acknowledge the usage of earlier terms in applicable settings.
For scientific researchers, university grants officers, and defense contractors, the bare abbreviation SI presents immediate communication hazards. SI also has other meanings, such as the International System of Units. Because of these conflicting meanings, research teams and acquisition personnel should spell out "Super Intelligence (SI)" upon first appearance in white papers, full proposals, technical reports, and public agency notices.
The substantive definition governing federal usage is codified in Section 3(a) of Executive Order 14434:
"For purposes of this order, and except where otherwise provided by law, the terms 'Super Intelligence' and 'SI' mean the technologies and systems encompassed by the term 'artificial intelligence' as defined in section 9401(3) of title 15, United States Code."
That statutory provision, 15 U.S.C. 9401(3), was enacted by Congress under the National Artificial Intelligence Initiative Act of 2020:
"The term 'artificial intelligence' means a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations or decisions influencing real or virtual environments."
Because Section 3(a) defines SI by reference to 15 U.S.C. 9401(3), federal Super Intelligence and artificial intelligence describe the same systems. In federal solicitations, the term encompasses machine learning models, neural network architectures, neuro-symbolic reasoning pipelines, computer vision systems, and automated decision logic. It establishes no requirement for artificial general intelligence, human equivalence, or autonomous consciousness.
This legal anchor creates a clear boundary between federal procurement language and computer science research concepts:
Federal Super Intelligence (two capitalized words, abbreviated as SI) is an administrative designation for current operational systems meeting the functional criteria of 15 U.S.C. 9401(3).
Research superintelligence (one compound word, commonly abbreviated as ASI for artificial superintelligence) is a hypothetical future capability that would surpass human experts across virtually every discipline, which does not exist today. Commercial laboratories, including Meta Superintelligence Labs, Microsoft's MAI Superintelligence Team, and Safe Superintelligence Inc., identify artificial superintelligence as an aspirational long-term research goal.
Federal research teams must also recognize the legal boundary set forth in Section 4(c) of Executive Order 14434. The text specifies that the order creates no right or benefit, substantive or procedural, enforceable at law or in equity by any party against the United States, its departments, agencies, officers, or any other person. Section 4(c) of EO 14434 establishes that the order creates no privately enforceable legal rights in procurement protests. Evaluators review BAA submissions strictly against the peer review factors published in the solicitation pursuant to Title 48 of the Code of Federal Regulations.
Related guides
- Super Intelligence in Federal Procurement: Acquisition Lifecycle and Agency PurchasingExecutive Order 14434 directs federal executive agencies to adopt Super Intelligence terminology across non-statutory...
- Commercial Software Vendor Branding After the Super Intelligence OrderExecutive Order 14434 directs federal agencies to adopt Super Intelligence terminology across non-statutory records,...
- Super Intelligence Catalog Updates for GSA Schedule Vendors: eMod GuideExecutive Order 14434 directs federal agencies to adopt Super Intelligence terminology across non-statutory documents...
- Super Intelligence in Government Contracts: Clause and SOW Analysis Under EO 14434Executive Order 14434 directs federal agencies to adopt Super Intelligence in non-statutory communications while...
- What the Super Intelligence Order Means for Federal Contractors: EO 14434 GuideExecutive Order 14434 directs federal agencies to adopt Super Intelligence in official communications while preserving...
- Super Intelligence Statutory Definition: 15 U.S.C. 9401 and the 2026 TimelineFederal agencies now use the term Super Intelligence to describe systems defined under 15 U.S.C. 9401(3). Under...
More on this subject: AI Agents: General Guides (172 guides)
Key Differences Between Broad Agency Announcements and Commercial RFPs
Broad Agency Announcements are often confused with standard Requests for Proposals (RFPs) issued under FAR Part 15 (Contracting by Negotiation). This confusion leads contractors to submit polished product brochures instead of scientific research methodologies. Understanding the regulatory mechanics of FAR 35.016 is essential for any organization responding to federal research solicitations.
A commercial RFP under FAR Part 15 begins with a defined government requirement, such as buying cloud infrastructure, accounting software, or maintenance support. In that environment, the agency publishes a rigid Statement of Work (SOW) or Performance Work Statement (PWS). Offerors submit proposals designed to meet that identical specification. A formal Source Selection Evaluation Board (SSEB) evaluates offerors against common Section M criteria, scoring vendors head-to-head to determine the best value tradeoff or lowest price technically acceptable bid.
A Broad Agency Announcement governed by FAR 35.016 operates under an entirely different procurement philosophy:
Scientific focus: BAAs are used exclusively for basic research, applied research, and that part of development not related to specific hardware or operational system procurements. Agencies use BAAs to advance scientific knowledge and explore high-risk, high-reward concepts.
Absence of a common statement of work: An agency issuing a BAA does not provide a single SOW. Instead, the announcement outlines broad research areas of interest, technical problem statements, or capability targets. Each offeror proposes their own unique scientific approach, hypothesis, and experimental design.
Independent peer review: Because submissions propose distinct technical approaches, proposals cannot be evaluated against each other. Evaluators assess each proposal independently against the criteria published in the BAA.
Flexible award instruments: FAR 35.016 governs BAAs that lead to procurement contracts; agencies also use announcements to award grants and cooperative agreements under their assistance authorities and, at DoD, other transactions under 10 U.S.C. 4021 and 4022.
The following comparison illustrates how Broad Agency Announcements diverge from standard negotiated procurements:
When agencies such as the Defense Advanced Research Projects Agency (DARPA), the Intelligence Advanced Research Projects Activity (IARPA), the Office of Naval Research (ONR), the Air Force Research Laboratory (AFRL) and the Advanced Research Projects Agency for Health (ARPA-H) publish BAAs on SI topics, they seek research advances rather than product demonstrations. Proposers must present mathematical models, empirical test protocols, and sound scientific reasoning rather than commercial software demonstrations.
Scientific Peer Review and Technical Merit Under FAR 35.016
Under FAR 35.016(e), the primary basis for selecting proposals for acceptance is technical merit, importance to agency programs, and fund availability, and cost realism and reasonableness are also considered to the extent appropriate. Agencies that adopted the Revolutionary FAR Overhaul deviation number this section 35.102.
Peer review panels consist of research scientists, military technical fellows, and university subject matter experts. Evaluators examine specific technical attributes:
Scientific novelty: Does the proposed approach introduce a distinct conceptual advancement beyond published literature, or does it merely repackage existing algorithmic baselines?
Technical feasibility: Is the mathematical and theoretical foundation sound, and does the research team demonstrate the laboratory capabilities necessary to execute the experiments?
Potential contribution: If the research succeeds, does it yield substantial capability enhancements for agency missions, such as autonomous perception in degraded environments or reliable automated theorem proving?
Qualifications of key personnel: Evaluators examine the principal investigator's publication record, prior experimental achievements, and technical depth in the relevant scientific domain.
Multi-Stage BAA Pipelines: From Concept White Papers to Full Proposals
Unlike commercial procurements where offerors produce multi-volume proposals immediately, BAAs typically employ a structured, multi-stage submission pipeline:
Stage 1: Executive summaries, concept abstracts, or research white papers. These submissions are brief, usually three to five pages. Offerors outline their scientific hypothesis, technical methodology, rough order of magnitude (ROM) cost, and project milestones.
Review and feedback: The cognizant agency Program Manager (PM) reviews submitted white papers against the BAA focus areas. The PM issues a formal response encouraging or discouraging a full proposal. In DARPA and IARPA programs, this filtering step saves research teams hundreds of hours of proposal preparation for approaches that do not align with program vision.
Stage 2: Full technical and cost proposals. Offerors who receive encouragement, or any offeror choosing to proceed under an open BAA, submit an in-depth technical volume, statement of work, milestone schedule, data rights assertions and detailed cost proposal.
Award negotiation: Selected proposals move into technical and financial negotiations, where the agency and offeror refine the statement of work, milestone deliverables, and intellectual property terms before final contract execution.
A 4-Step Framework for Responding to Super Intelligence Solicitations
Navigating federal research solicitations under Executive Order 14434 requires research institutions and defense laboratories to adapt their proposal preparation methods. Proposal teams must align with new administrative terminology while preserving the scientific rigor expected by federal peer review panels.
Teams responding to Super Intelligence BAAs should apply this 4-step framework:
Analyze technical topic descriptions and agency problem statements: Examine the BAA technical areas to determine whether the agency requests fundamental basic research (6.1 funding), applied research (6.2 funding), or advanced technology development (6.3 funding). Map the agency's stated capability objectives against your institutional research strengths, noting whether the solicitation references specific federal risk frameworks or algorithmic benchmarks.
Align scientific white papers with federal SI terminology and 15 U.S.C. 9401(3): Where the BAA uses "Super Intelligence (SI)," mirror it in the executive summary, problem statement and title page; the order directs agencies, not proposers, so this is about responsiveness, not a requirement. Ground all technical claims directly in the functional definition of 15 U.S.C. 9401(3), explicitly defining your system as a machine-based system performing predictive, analytic, or decision-support tasks.
Structure empirical benchmarks and technical readiness levels: Define concrete, falsifiable evaluation metrics for each research phase. Specify quantitative baselines, such as accuracy thresholds, inference latency constraints, sample efficiency ratios, or adversarial perturbation tolerances. Map project milestones to Technology Readiness Levels (TRLs) to show a structured progression from laboratory proof of concept to field validation.
Establish collaborative research workspaces and deliverable repositories: Scientific R&D involves multi-institutional consortia encompassing prime contractors, university research labs, and specialized commercial subcontractors. Deploy a unified workspace to manage proposal drafts, literature reviews, mathematical proofs, experimental datasets, and sub-award documentation under granular access controls and version tracking.
Applying this framework prevents the common error of treating federal Super Intelligence as an exotic, speculative technology, grounding the proposal firmly in verifiable engineering principles.
Handling Academic Citations and Prior Research Under Section 2(b)
A frequent point of friction in federal proposal writing is reconciling new administrative terminology with decades of established scientific literature. Academic papers, open-source repositories, and foundational benchmarks use the terms "artificial intelligence," "deep learning," and "machine learning."
Section 2(b) of Executive Order 14434 reads:
"Nothing in this section requires the alteration of previously issued regulations, Presidential actions, contracts, grants, or other historical documents."
Research teams must maintain strict fidelity to scientific source materials. When citing prior publications, research grants, or academic conference proceedings in BAA white papers, authors should preserve the original titles and terminology used by the authors. For instance, citing the "National Artificial Intelligence Initiative Act" or a research paper titled "Reinforcement Learning in Multi-Agent Environments" requires zero alteration.
To prevent confusion among evaluators, proposal teams can include a brief terminology note in the introduction:
"Note on terminology: This proposal uses 'Super Intelligence (SI)', the term used in this announcement under Executive Order 14434, to mean artificial intelligence as defined in 15 U.S.C. 9401(3). Cited publications, grants and datasets keep their original published titles."
This practice demonstrates procedural awareness while upholding scientific and bibliographical accuracy.
Structuring Empirical Milestones and Technical Readiness Levels
Federal R&D solicitations evaluate whether a proposer can translate abstract mathematical hypotheses into measurable technical deliverables. Program Managers construct BAAs around defined program phases, each ending in a rigorous go/no-go milestone review.
Proposers can structure their technical approach across Technology Readiness Levels:
Phase 1 (TRL 2 to 3): Analytical formulation and experimental proof-of-concept. Focus on mathematical derivations, baseline algorithm training, and synthetic benchmark validation in controlled laboratory settings.
Phase 2 (TRL 4 to 5): Component validation in representative environments. Integrate algorithmic modules with simulated sensor streams, test resilience against out-of-distribution inputs, and measure inference throughput under realistic hardware constraints.
Phase 3 (TRL 6): System demonstration in operational environments. Deploy containerized software modules onto government testbeds or tactical hardware prototypes, validating real-time predictive performance during live exercises.
Each phase must include explicit quantitative metrics. Rather than proposing "to improve model accuracy," a compliant proposal commits to exceeding published baseline accuracy metrics on the government-provided evaluation corpus while satisfying specified operational inference latency limits on edge compute hardware. These concrete commitments provide peer reviewers with the empirical justification required to award top technical ratings.
Managing BAA Research Deliverables, Consortia, and Audit Trails
Federal research initiatives under Broad Agency Announcements rarely occur in isolation. Prime defense contractors partner with leading research universities, national laboratories, and specialized software startups to build interdisciplinary research teams. Managing collaborative research across institutional boundaries creates complex document governance challenges.
Throughout the lifecycle of a BAA award, consortia generate hundreds of technical assets: multi-stage white papers, mathematical proofs, experimental datasets, quarterly technical progress reports, and formal Contract Data Requirements Lists (CDRLs). University co-investigators must review draft submissions without gaining unauthorized access to proprietary corporate labor rates or indirect cost structures. Similarly, government Program Managers require clear visibility into deliverable revisions and validation data.
When proposal and research files are scattered across personal drives, consumer cloud folders, and email threads, teams experience configuration drift, missed milestone deadlines, and version collisions that threaten project execution.
Fast.io workspaces provide secure cloud infrastructure designed for distributed research teams and their AI agents. Organizations configure dedicated, org-owned workspaces for specific BAA topic areas or multi-year research awards, establishing a centralized repository for all project documentation, data models, and contractual deliverables.
Key workspace capabilities for federal research operations include:
Granular access controls: Permissions are set per organization, workspace and share; a share gives an outside party access to only the files or folder placed in it. Prime contractors can create shares for academic partners to upload white papers and benchmark data while isolating proprietary cost volumes, intellectual property disclosures, and executive communications.
Per-file version history: Every research paper, dataset schema, and technical volume maintains a complete, restorable revision history. When research teams incorporate peer review feedback or update experimental results across project milestones, version history preserves every iteration, preventing accidental overwrites.
Detailed activity log: Fastio keeps a detailed activity log of file uploads, changes, and permission and membership changes, so project directors and compliance officers can see who changed which file and when.
Metadata Views for solicitation extraction: BAA solicitation packages, broad topic calls, and program amendments frequently exceed hundreds of pages of complex specifications. Metadata Views convert unstructured solicitation documents and research PDFs into structured, queryable data. Teams define extraction parameters in natural language, and the system extracts submission deadlines, technical topic IDs, CDRL deliverables, and required reporting intervals into a sortable spreadsheet view without manual data entry.
Integrated workspace e-signature: Teaming agreements, non-disclosure covenants, and university sub-award commitment letters can be routed for execution directly within the workspace using Fast.io e-signature. The platform supports sequential or parallel signer routing, optional one-time passcodes, and a tamper-evident audit certificate filed automatically into the workspace upon execution.
Fast.io provides transparent subscription tiers on Fast.io pricing. Organizations begin evaluating the platform through a 14-day trial on monthly plans, which requires a credit card.
As research consortia expand across multiple BAA programs, teams can add additional storage and AI work credits to match operational requirements.
Coordinate Federal Research Proposals and White Papers in One Workspace
Organize multi-stage BAA white papers, research datasets, and technical amendments across secure team workspaces. Monthly plans include a 14-day trial; a card is required.
The November 2026 Legislative Timeline and Future R&D Solicitations
Federal research institutions and contractors must navigate immediate BAA solicitations while preparing for longer-term statutory and regulatory developments. Executive Order 14434 establishes an administrative timetable that will influence future research announcements.
Section 3(b) of Executive Order 14434 instructs the Assistant to the President for Science and Technology, who leads the White House Office of Science and Technology Policy (OSTP), to submit proposed legislative language establishing a federal definition of "Super Intelligence" and "SI" to the President within 60 days of the signing date. Because the order was signed on September 29, 2026, this 60-day milestone falls on November 28, 2026.
Federal contractors and researchers must understand the legal status of this milestone:
As of October 2026, no statutory definition of Super Intelligence exists in federal law.
The November 28, 2026 milestone represents an executive recommendation submitted to the President, not an enacted statute. An executive order cannot amend federal statutory law; only an Act of Congress signed by the President can update the United States Code.
What the APST's legislative proposal will recommend, and whether Congress will introduce or enact conforming statutory language, remains unknown as of October 2026.
Until later Presidential action or an Act of Congress changes it, SI under the order means what 15 U.S.C. 9401(3) defines as artificial intelligence, and each solicitation's own definitions still govern that solicitation.
Following the November 28, 2026 report, research organizations should anticipate three potential governance developments:
Agency-level BAA topic updates: Research agencies such as DARPA, ARPA-H and the Office of Naval Research may use "Super Intelligence (SI)" in new BAA amendments, program titles and announcements published on SAM.gov.
Federal Acquisition Regulation (FAR) Council rulemaking: If the administration chooses to formalize standard Super Intelligence contract clauses across federal procurements, the FAR Council would draft proposed amendments to Title 48 of the Code of Federal Regulations. This process requires publishing proposed rules in the Federal Register, conducting a public notice-and-comment period, and issuing final rules before standard clauses appear in federal contracts.
Congressional legislative deliberations: If draft legislation is transmitted to Capitol Hill, committees of jurisdiction, including the House Committee on Science, Space, and Technology and the Senate Committee on Commerce, Science, and Transportation, would evaluate the bill, conduct hearings, and debate potential statutory revisions.
To maintain proposal readiness during this transitional period, research organizations should implement three operational practices:
Monitor agency BAA portals: Track topic descriptions and program updates on SAM.gov and agency-specific portals (such as DARPA's BAA portal). Always mirror the agency's primary terminology across white paper narratives and technical proposal volumes.
Protect historical research integrity: Continue citing published academic literature, historical grant awards, and baseline algorithms using their original published titles, relying on the Section 2(b) savings clause.
Engage Contracting Officers early: When reviewing draft BAAs or attending agency Proposers' Days, submit formal clarification requests if terminology instructions appear ambiguous across technical areas or CDRL reporting guidelines.
Sources
References used to verify factual claims in this guide.
-
Replacing artificial intelligence with Super Intelligence across federal agency communications is likely to produce administrative and procedural adjustments rather than substantive statutory changes in the near term.
-
Under 15 U.S.C. 9401(3), artificial intelligence is defined as a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations or decisions influencing real or virtual environments.
Frequently Asked Questions
How do federal agencies use Super Intelligence in Broad Agency Announcements?
Federal agencies use 'Super Intelligence' and the abbreviation 'SI' in Broad Agency Announcements pursuant to Executive Order 14434, which directs executive agencies to substitute these terms for 'Artificial Intelligence' and 'AI' in non-statutory documents. Under Section 3(a) of the order, federal Super Intelligence is legally anchored to the statutory definition of artificial intelligence in 15 U.S.C. 9401(3).
Does EO 14434 change technical evaluation in DARPA BAAs?
No. Technical evaluation in DARPA BAAs continues to follow the scientific peer review process established in FAR 35.016. Evaluators assess submissions based on scientific merit, technical feasibility, potential contribution to agency missions, and team qualifications. Because Section 3(a) references 15 U.S.C. 9401(3), evaluators review the exact same underlying technologies, such as machine learning and automated decision logic, regardless of administrative labeling.
How should research teams cite AI systems in federal BAA white papers?
Research teams should mirror the solicitation's terminology by using 'Super Intelligence (SI)' in the executive summary, project title, and proposal narrative while defining technical architectures pursuant to 15 U.S.C. 9401(3). Prior peer-reviewed literature, open-source models, and historical grants should retain their original published titles under the savings clause in Section 2(b) of Executive Order 14434.
What is the difference between a Broad Agency Announcement and a FAR Part 15 RFP?
A FAR Part 15 RFP buys specific products or services against a rigid Statement of Work, evaluating offerors head-to-head through a Source Selection Evaluation Board. A Broad Agency Announcement under FAR 35.016 solicits proposals for basic and applied research across broad topic areas, evaluating each submission independently on its individual scientific merit through peer review without a common Statement of Work.
Does Section 2(b) of Executive Order 14434 protect historical research grants?
Section 2(b) states that nothing in that section requires the alteration of previously issued regulations, Presidential actions, contracts, grants, or other historical documents, so existing research grants keep their wording.
When will Congress enact a statutory definition of Super Intelligence?
As of October 2026, no statutory definition of Super Intelligence exists in federal law. Section 3(b) of Executive Order 14434 requires the Assistant to the President for Science and Technology to submit proposed legislative language to the President by November 28, 2026. That submission is an executive recommendation to the President, not an enacted statute; any formal definition requires congressional legislation.
Can a disappointed proposer protest a BAA award based on Executive Order 14434?
The order itself does not supply grounds for one: Section 4(c) states that it creates no right or benefit, substantive or procedural, enforceable at law or in equity against the United States or its agencies. Any bid protest before the Government Accountability Office (GAO) or U.S. Court of Federal Claims must be grounded in procurement statutes, the FAR, or the published evaluation criteria of the BAA.
Related Resources
Coordinate Federal Research Proposals and White Papers in One Workspace
Organize multi-stage BAA white papers, research datasets, and technical amendments across secure team workspaces. Monthly plans include a 14-day trial; a card is required.