Start here — a map of the pack

This guide provides an overview of the Institutional AI Readiness Pack, explaining what each instrument is designed to do and how to get started.

The Institutional AI Readiness Pack is a university-wide assessment and implementation toolkit designed for the responsible use of AI in academic research. It spans practice, people, policy, systems, procurement, data, disclosure, and oversight. It serves as a practical companion to Responsible AI in Academic Research: A Competency Framework for Research Training, which defines five core dimensions of institutional readiness along with the underlying capabilities needed to support them.

This pack translates that framework into practical instruments that your university can use to build an evidence-based picture of how AI is actually used and governed across your research environment. It also gives you ways to track improvements over time. By doing this, it connects high-level institutional policy and strategic priorities directly with the everyday practices, skills, and experiences of researchers and graduate students.

Browse the full online instrument catalogue to read or download every resource in HTML, Word, PDF, and spreadsheet formats.

What's in it

Start with the main document. The Institutional AI Readiness Pack is the foundation for everything here. It explains how the report's framework is used across the pack, walks you through how to run an assessment and interpret your results, and illustrates the entire process using a fictional university from its initial assessment to its first concrete actions. I recommend reading it once before you begin, and keeping it on hand as a reference while you run the assessment.

Thirteen instruments across four groups.

A companion scoring workbook. I also provide a spreadsheet that handles all the calculations. Once you enter your A1 self-assessment responses, the workbook automatically generates your institutional profile, highlights the specific axis holding each dimension back, and identifies the single dimension that is constraining your overall progress.

Three things to know before you start

The assessment uses a triangulated design. Research leadership completes the maturity self-assessment (A1), doctoral students complete the student survey (A2), and advisors and supervisors complete the supervisor survey (A3). The surveys for students and supervisors include parallel questions that ask the anonymous cohorts about the same institutional artifacts and practices, alongside questions specific to each role. You will want to account for differences in response scales before comparing their answers directly. Comparing these different perspectives is essential because the gap between what leadership can document and what graduate students and supervisors report experiencing is where the most valuable insights emerge. An institution that only surveys central administrators will only confirm what it already assumes.

A1 requires documentary evidence for every answer. Each A1 indicator asks whether a specific, verifiable artifact exists, such as an approved policy clause, a designated role with a current occupant, an active software register, or a verified training record. If the evidence establishes that the requirement is not met, record No. If the evidence is unavailable or unresolved, leave the item unanswered rather than treating the blank as a documented No. A2 and A3 collect respondent evidence and do not require a documentary artifact for every response.

I intentionally avoid reducing your institution to a single overall score. Instead, the assessment produces individual maturity levels across five distinct dimensions and identifies the specific bottleneck or constraint that is holding your institution back. Combining everything into a single average score would allow a well-written policy document to mask an untrained student cohort or an unmonitored IT environment. In practice, that kind of averaging is the most common way universities mislead themselves about their actual AI readiness.

Where to begin

1. Read the briefing tailored to your role (B1 to B5). This takes about ten minutes and outlines the primary responsibilities and strategic decisions that sit with you.
2. Convene your working group to complete the maturity self-assessment (A1). This typically takes half a day. Your group should include representatives from the research office, research integrity, the graduate school, central IT, the library, and at least two active researchers. This breadth is necessary because no single office holds all the required documentary evidence.
3. Field the student and supervisor surveys (A2 and A3). You can run these surveys while A1 is being completed. Invite the full eligible cohorts, report the achieved response rates, and avoid describing respondents as representative without comparing them with the eligible populations.
4. Work through the gap analysis and action ladder (A4). This instrument organizes required improvements into priority work packages toward target rungs. For each package, it identifies a concrete action, a named owner, and the artifact used in the next A1 rescore.

In my experience, most universities can complete the full assessment cycle over one academic quarter, requiring roughly twenty-two to twenty-five person-days of effort across the working group, alongside the time contributed by survey respondents.

Adapt all of it

Everything here is licensed CC BY 4.0. You may rewrite it, shorten it, translate it, rename it, and adopt it as your own institutional document (including commercially). You do not need to ask permission, and you do not need to notify Instats. Attribution is the only condition: Adapted from The Institutional AI Readiness Pack (Instats, 2026), CC BY 4.0.

Every university operates in a distinct regulatory and operational environment, so a toolkit that you cannot legally customize would not be very helpful. These instruments are designed to be adapted to fit your institution's specific needs.

Questions: support@instats.org · instats.org

Michael J. Zyphur, PhD · Professor and Director, Instats · instats.org · support@instats.org