Government AI Governance & Research

Agencies are already letting AI decide. The evidence that oversight is real still isn't there.

Automation moved past the pilot stage. AI now touches benefits determinations, fraud flags, procurement scoring, and case triage — decisions that used to sit with a person. What hasn't kept pace is proof: that the human still in the loop is catching what matters, and that the decision can be reconstructed after the fact.

The GIAG research program gives government AI leaders an evidence base for adoption decisions, built from practitioner data, independent of vendor marketing.

Explore Our Research Join a Research Initiative Subscribe to Newsletter

Why It Matters

Governance without measurement is aspiration without accountability.

Federal mandates, OMB guidance, and NIST frameworks have established the what of AI governance. ThinkCapital's research program is focused on the harder questions: the how, the how much, and the how do we know. We work at the intersection of measurement science, organizational behavior, and government IT, producing frameworks practitioners can actually use.

Four questions now separate real governance from paperwork:

  • Program & AI leads — If a reviewer signs off on an AI-assisted decision in seconds, is that human oversight, or a signature?
  • Legal, IG & oversight teams — If an AI-assisted decision were challenged tomorrow, could your agency reconstruct exactly how it was reached, and who was accountable for it?
  • Governance & compliance teams — Is your AI governance framework built on evidence and measurement, or on policy compliance and good intentions?
  • Budget & investment decision-makers — Do you have a way to know whether the AI is deciding better than what it replaced, or are you still taking the vendor's word for it?
View Research Program → Reference Body of Knowledge
GIAG — A ThinkCapital Research Program

ThinkCapital Research Initiatives

Active & Forming Research Streams

These are the current focus areas driving ThinkCapital's research agenda. Visit individual initiative pages to read full descriptions, track progress, or express your interest in participating.

● Active

NIST AI RMF Implementation in Government Contexts

Examining how the NIST AI Risk Management Framework is actually being adopted, or not, across federal and state agencies, and what factors predict successful operationalization.

Read Initiative → Express Interest
● Active

Meaningful Human Oversight: From Requirement to Practice

What does meaningful human oversight of AI actually look like in operational government settings? We are developing measurable indicators and maturity criteria.

Read Initiative → Express Interest
◌ Forming

AI Adoption Thresholds & Organizational Tipping Points

Applying Schelling-Granovetter threshold models to explain why identical AI initiatives succeed in some agencies and stall in others, and how to move the needle.

Read Initiative → Express Interest
◌ Forming

AI Productivity Measurement for Government Missions

Developing empirical methods for measuring whether AI tools are delivering measurable productivity improvements in government mission contexts, beyond technology metrics alone.

Read Initiative → Express Interest

Advisory Services

Need help now, before the research is finished?

ThinkCapital also provides direct advisory services to government agencies and their contractors navigating AI governance, technology investment decisions, and organizational readiness. Our work is grounded in the same measurement discipline as our research.

See Our Services Who We Are

"The agencies that will lead in AI are not the ones that move fastest — they are the ones that measure best."

ThinkCapital Research Perspective

Get Involved

Questions, comments, or ready to engage?

Whether you want to follow our research, participate in an initiative, or simply ask a question, the Engage page is where conversations begin.

Go to the Engage Page →