︎
READY STATE LABS

Summary


Ready State Labs (RSL) is a non-profit do-tank dedicated to helping the U.S. federal government keep pace with rapidly advancing AI. As artificial intelligence transforms society, RSL functions as an experimental and entrepreneurial lab made up of policy, government modernization, and AI experts working to equip the federal government with the expertise and capacity needed to govern and deploy AI effectively. RSL helps policymakers and implementers understand emerging AI capabilities, transform the institutions responsible for governing and adopting them, and develop practical tools that improve how government serves the American people.

Frontier AI is advancing faster than the public institutions responsible for overseeing it. Today's AI policy field is crowded with debates about safety, national security, competition, and regulation, but few address whether government has the knowledge, infrastructure, and capacity to keep pace with the technology it is being asked to govern.

America’s federal government currently struggles to:
    • Develop informed policy. Policymakers today lack the information and expertise needed to understand emerging AI capabilities before they become urgent policy challenges.
    • Establish the technical foundations for adoption. Agencies and congressional institutions lack shared standards, modern data practices, evaluation frameworks, and clear technical requirements that would allow AI systems to be adopted securely, responsibly, and across institutional boundaries.
    • Translate emerging capabilities into practical use. Congress and agencies often lack the product capacity, implementation support, and access to real-world testing needed to turn promising AI capabilities into tools that work within government. 

Left unaddressed, these gaps will keep government reactive, making consequential decisions about AI without the capacity to shape its use. RSL is a cross-partisan organization built to close them: helping Congress and executive branch agencies make informed policy, establish the technical foundations for effective AI adoption, and build the practical tools government needs to lead in the AI era.


Methods for Impact


RSL’s core insight is that government readiness cannot be built through isolated policy, infrastructure, or product interventions. These layers are interdependent: policy creates the authorities, incentives, standards, and resources needed to modernize government infrastructure; stronger infrastructure makes it possible to develop and deploy useful products; and real-world product development reveals the policy and infrastructure changes still required.



Policy enables infrastructure, and infrastructure enables product development. Lessons from product development and use feed back into policy.

  1. Policy

    Set the rules and direction

    • Strategy & education
    • Standards & incentives
    • Implementation guidance

  2. Infrastructure

    Build the enabling systems

    • Data standards
    • Interoperable systems
    • Technical capacity

  3. Product

    Turn capacity into practical tools

    • Use cases & pilots
    • Workflow tools
    • Adoption & learning


Policy: Equip policymakers with technical expertise through research, briefings, training, and direct support on emerging AI capabilities and their implications. This work will include:
    • Supporting policymakers as they evaluate the laws, rules, standards, funding mechanisms, and governance structures needed to enable infrastructure modernization and responsible AI adoption; and
    • Convening policymakers, technical experts, civil society, government practitioners, and AI lab advisors around urgent governance and implementation questions.

Infrastructure: Identify and address the shared technical and institutional prerequisites that an AI-ready government requires. This work will include:
    • Improving priority government data systems so they are usable for appropriate AI-enabled tools;
    • Developing standards for metadata, MCPs, interoperability, and machine-understandability;
    • Supporting procurement and evaluation frameworks for AI systems;
    • Building internal technical governance practices for testing, monitoring, and auditing AI tools; and
    • Helping public institutions move from one-off pilots to reusable infrastructure that can support many AI use cases over time.

Product: Research, develop, and test practical products designed around real legislative and executive branch workflows. The key advantage of the product layer will be its ability to establish trusted partnerships on both sides of the development process, with congressional offices and agencies willing to serve as early users, and with AI labs able to contribute technical talent, compute, tools, and expertise. RSL serves as the trusted intermediary responsible for translating government needs, coordinating partnerships, and ensuring that resulting products serve public-sector priorities. Initial focus areas will include:
    • Developing open-source AI tools for legislative and executive branch use cases and cataloging existing use cases;
    • Testing AI applications against real government workflows, data constraints, and operational needs;
    • Creating external evaluation methods for AI products used in public-sector contexts;
    • Maintaining a core research and engineering team that can rapidly prototype tools in partnership with public institutions; and
    • Funding full-time FDEs to support the last-mile execution of public-sector policy priorities.


Strategic Priorities


RSL’s initial work, chosen to meet AI’s rapidly advancing capabilities, will focus on three strategic priorities:

Applied AI Research
RSL will conduct applied research on emerging AI capabilities, their implications for government, and the practical conditions under which they should be adopted (e.g. AI auditing or security tools). This work will translate technical developments into practical evidence, briefings, pilots, and recommendations that help policymakers and implementers understand where AI can improve government workflows, where adoption is premature or risky, and what governance questions require action.

AI Security and Cyber Resilience
As recent agent-orchestrated attacks have revealed, agents can reason across codebases, operate tools autonomously, and surface exploitable vulnerabilities faster than most government institutions can detect or respond to them, an asymmetry that will only widen with each model generation. RSL will support the government in building what it needs to anticipate emerging AI-enabled threats and build practical security capacity.

AI-Ready and Agent-Ready Government
RSL will help government identify and address the shared technical and institutional prerequisites that enable a flexible and resilient AI-ready government. This will include developing the standards, frameworks, and technical architecture required to build internal AI tools and make public information usable by external models and agents.

RSL will work across these three strategic priorities, using its three layered methods for impact. 
The matrix below illustrates how these areas of work intersect.

Illustrative activities across RSL's three strategic priorities and three areas of work.
OUR PRIORITIES PolicySet the rules and direction InfrastructureBuild the enabling systems ProductTurn capacity into practical tools
Applied AI
Research

Translate the frontier

Capability assessments, technical briefings, and recommendations for AI adoption and oversight.

Research in practice

Agency partnerships, embedded technical teams, and shared environments for applied research and policy pilots.

Test what works

Evaluations, benchmarks, and prototypes that show what emerging AI can reliably do in government.
AI Security &
Cyber Resilience

Stronger safeguards

Recommended AI security requirements, procurement criteria, and cross-agency coordination frameworks.

Defensive capacity

Security testbeds, red-team programs, and incident-reporting processes to evaluate and respond to emerging threats.

Tools for defenders

Open-source tools for vulnerability detection, security monitoring, auditing, and AI-enabled defense.
AI-Ready
Government

Standards for readiness

AI- and agent-ready data standards, access rules, and proposals to strengthen technical delivery capacity.

Shared foundations

Structured, permissioned data environments, shared architecture, and internal teams to build and maintain systems.

Better everyday workflows

Reusable APIs and agent connectors, internal knowledge tools, and open-source congressional and agency applications.
ILLUSTRATIVE ACTIVITIES

Priorities may evolve as AI capabilities and government needs change.



Why Now?


Frontier AI capabilities are advancing on a timeline measured in months, not years, forcing the federal government to make decisions with global consequences at a speed its institutions were not designed to match. The next six to eighteen months will bring a series of unusually consequential policy, funding, and implementation decisions that will determine whether government builds the capacity it needs to keep pace or locks in another generation of technical debt. 

The most immediate opportunities, all of which require work to begin within the next six months, include:
    • Supporting the incoming 120th Congress with recommendations for congressional rules modernization, emerging AI policy packages, and AI-related provisions in the NDAA and appropriations process.
    • Preparing policymakers for increasingly capable models, from policy recommendations and legislative language, to oversight frameworks, evaluation standards, and implementation guidance.
    • Turning federal AI-ready data priorities into reality, including developing concrete standards, agency plans, and 2027 appropriations requests in response to the National Policy Framework for AI’s recent call for accessible AI-ready federal datasets for use in training AI systems.
    • Integrating emerging technologies into government workflows as they mature, including rapidly developing AI security, evaluation, and auditing tools.