What is this course about?
Pre-publication research is the most sensitive data on a campus and the least protected by typical AI policy. This course covers where AI genuinely accelerates research administration — solicitation parsing, compliance checklists, effort reporting — and where sponsor terms, export control, and IRB obligations make autonomous AI use unacceptable.
Who is this course for?
- Sponsored programs and research administration staff
- IRB administrators and chairs
- Research compliance and export control officers
- Faculty with substantial grant portfolios
What do I need before starting?
- Familiarity with the sponsored research lifecycle at your institution
- No technical background required
What will I be able to do afterwards?
- Classify research data by AI-eligibility under sponsor and export control terms
- Use AI to parse solicitations and build compliance checklists reliably
- Identify where AI may assist IRB review and where it must not decide
- Recognize export control and ITAR exposure in research computing
- Deploy a solicitation analysis agent inside the institutional boundary
What does each module cover?
Why is pre-publication research the hardest data to protect?
40 minThe asymmetry: enormous value, weak default protection, and researchers who will use consumer tools regardless.
Objectives
- Characterize the risk profile of unpublished research
- Identify where researchers are already using unapproved tools
- Frame the institutional response as enablement rather than prohibition
Topics
Activity. Survey your own unit's actual AI tool use and compare it with approved-tool policy.
How do you parse a solicitation reliably?
55 minTurning a 90-page funding opportunity into a verified compliance checklist.
Objectives
- Extract requirements from a complex solicitation
- Verify extraction completeness against the source
- Generate a checklist a PI will actually use
Topics
Activity. Parse a real solicitation and verify every extracted requirement against the document.
Where can AI assist IRB review — and where must it not?
50 minProtocol review support that speeds administration without displacing the board's judgment.
Objectives
- Identify administrative IRB tasks suitable for AI assistance
- State clearly where board judgment cannot be delegated
- Design completeness checking that reduces revision cycles
Topics
Activity. Build a protocol completeness checker and test it against previously returned submissions.
What do sponsor terms say about AI tools?
45 minThe clauses in federal and foundation awards that restrict AI use and data handling.
Objectives
- Locate AI and data handling restrictions in award terms
- Distinguish agency-level policy from award-specific terms
- Build a term-tracking process across a portfolio
Topics
Activity. Review three real award agreements for AI and data handling restrictions.
When does research computing trigger export control?
50 minITAR, EAR, and controlled unclassified information — where an AI tool becomes a deemed export.
Objectives
- Recognize export-controlled research in your portfolio
- Understand deemed export in the context of cloud AI services
- Specify the computing environment controlled work requires
Topics
Activity. Assess three projects for export control exposure and specify the computing environment each needs.
How do you use AI in post-award compliance?
45 minEffort reporting, subrecipient monitoring, and closeout — high-volume administrative work with clear rules.
Objectives
- Apply AI to effort reporting review
- Support subrecipient monitoring at portfolio scale
- Prepare closeout documentation efficiently
Topics
Activity. Build a subrecipient monitoring workflow and test it against a real portfolio sample.
How do you screen conflicts of interest at scale?
45 minCOI screening across disclosures, publications, and sponsor relationships.
Objectives
- Design COI screening across multiple data sources
- Handle entity resolution across name variants and affiliations
- Route findings for human determination
Topics
Activity. Run a COI screen against a synthetic disclosure set and evaluate precision and recall.
Deploying a solicitation agent inside the boundary
60 minThe hands-on module: an analysis agent running entirely within institutional infrastructure.
Objectives
- Deploy an agent with no external data egress
- Verify the boundary with network-level testing
- Document the deployment for a compliance review
Topics
Activity. Deploy the agent and prove with network testing that no research content leaves the boundary.
What is the capstone project?
Research administration AI plan with an export control assessment
Produce a plan for AI across your research administration lifecycle, including data classification by sponsor and export control terms, the IRB assistance boundary, and a verified in-boundary deployment for at least one workflow.
Deliverable: A plan plus a deployment with documented egress verification.
How are learners assessed?
- Solicitation extraction verified for completeness against the source document
- Export control classification exercise scored against a reference determination
- Egress verification evidence reviewed for technical sufficiency
What ships with the course?
Facilitator guide
Session-by-session running order, discussion prompts, and the questions that reliably derail a room.
Learner workbook
Exercises, checklists, and the templates each module's activity produces.
Hands-on lab environment
A sandboxed ibl.ai deployment so exercises run against real agents, not screenshots.
Assessment bank
Scenario questions and rubric criteria mapped to each stated learning outcome.
Source bibliography
Every primary regulation and standard cited on this page, linked and dated.
Which AI agents does this course use?
The hands-on modules run against agents already deployable on the ibl.ai platform for higher education.
Where does the course material come from?
Every module is grounded in primary sources — the regulation, standard, or research itself, not a summary of it. Each was resolved at authoring time.
- National Science Foundation
NSF
Agency policy on AI use and data handling in sponsored research.
- ClinicalTrials.gov
NIH
Registration and reporting obligations relevant to IRB and post-award modules.
- Directorate of Defense Trade Controls
U.S. Department of State
Authoritative source for ITAR obligations covered in Module 5.
- NIST SP 800-171 Rev. 3
NIST
Control requirements for controlled unclassified information in nonfederal systems.
Delivery notes
Binding guidance for anyone preparing and delivering this course.
- Module 5 requires genuine export control expertise. Have the institution's empowered official review it — a general compliance overview is not adequate for ITAR content and getting it wrong has criminal exposure.
- The Module 1 shadow-AI survey is uncomfortable and essential. Run it anonymously; the gap between policy and practice is the reason the course exists.
- Module 8's egress verification must be a real network test with packet-level evidence, not a configuration screenshot. Compliance reviewers will ask for the former.
- Solicitation examples should span agencies — NSF, NIH, and a DoD opportunity have materially different structures and the extraction approach differs.
- Do not let the IRB module suggest AI can make determinations. Every artifact should reinforce that the board decides and AI prepares.
Why run AI training on a platform you own?
You own the course, not a licence to it
Course content, learner data, and the platform run inside your perimeter — you own all the code and the data.
Model-agnostic delivery
Run the course's AI components on any LLM — Claude, GPT, Llama, Gemini, Command — and switch anytime.
No per-seat training licences
Usage-based or self-hosted, so cost tracks actual use rather than headcount.
Deploy anywhere
Cloud, private VPC, on-premise, or fully air-gapped — including for cohorts that cannot use public AI tools.
Frequently asked questions
What does the AI for Research Administration: Grants, IRB, and Compliance course cover?
Pre-publication research is the most sensitive data on a campus and the least protected by typical AI policy. This course covers where AI genuinely accelerates research administration — solicitation parsing, compliance checklists, effort reporting — and where sponsor terms, export control, and IRB obligations make autonomous AI use unacceptable. It runs 6 hours across 8 modules across 8 modules, at intermediate level, and closes with a capstone: Research administration AI plan with an export control assessment.
Who should take AI for Research Administration: Grants, IRB, and Compliance?
It is written for Sponsored programs and research administration staff, IRB administrators and chairs, Research compliance and export control officers, Faculty with substantial grant portfolios. Prerequisites: Familiarity with the sponsored research lifecycle at your institution; No technical background required.
Can we run this course on our own infrastructure?
Yes. ibl.ai is model-agnostic and deploy-anywhere — cloud, private VPC, on-premise, or fully air-gapped — and you own all the code and the data. Cohort data, submissions, and any material learners upload stay inside your perimeter, which matters for higher education teams that cannot send work to a public AI tool.
How do we get access to AI for Research Administration: Grants, IRB, and Compliance?
Request access and we will set it up for your cohort — hosted by ibl.ai, or running against your own deployment. Tell us the group size and timing you need, and whether it should run inside your own perimeter.
How much does AI training for higher education cost on ibl.ai?
There is no per-seat pricing — you pay for usage or self-host and pay only for the infrastructure, so a 5,000-person rollout does not cost 5,000 licences. 1.6M+ users across 400+ organizations run the platform this way, including NVIDIA, MIT, and Syracuse University.