Research initiative · Protocol v0.1
Pakistan AI & Futures Readiness Index 2027
PAFRI is being designed to measure whether people and higher-education institutions are prepared not only to use today's AI tools, but to keep learning, governing and making responsible decisions as the technology and the future of work keep changing.
Status: instrument development. No national benchmark results have been collected or published yet. National fieldwork will begin only after expert review, cognitive testing, pilot validation and ethics approval.
The research question
Readiness is more than AI adoption
Recent Pakistani research has already measured important aspects of student AI literacy and institutional AI readiness. PAFRI is designed to test a broader model that adds responsible judgement, learning agility, Futures Literacy and anticipatory institutional capacity.
People Readiness Index
A capability-focused index for students, faculty and higher-education leaders. Tool-use frequency is recorded separately and is not treated as competence.
Institutional Readiness Index
A separate institutional index covering governance, curriculum, access, infrastructure and the capacity to experiment and adapt under uncertainty.
People readiness
Four capabilities that should survive tool churn
AI Literacy & Critical Evaluation
Understanding what AI can and cannot reliably do, verifying claims, recognising failure modes and choosing when AI is appropriate.
Responsible AI Judgement & Human Agency
Privacy, accountability, meaningful human review, fairness, inclusion and the ability to keep responsibility visible.
Learning Agility & Human Capabilities
Learning new tools, transferring knowledge across systems and preserving judgement, reasoning and domain expertise as technology changes.
Futures Literacy & Foresight
Surfacing assumptions, considering multiple plausible futures, noticing signals of change and using scenarios to improve present decisions.
Institutional readiness
Four conditions for responsible, future-ready adoption
Governance & Accountability
Clear AI guidance, named accountability, review and appeal mechanisms, and policies that evolve with tools and risks.
Curriculum & Capability Development
AI literacy beyond computing disciplines, faculty development, assessment redesign and future-oriented capability planning.
Inclusion, Access & Infrastructure
Connectivity, approved tools, privacy and security, accessibility, affordability and workable alternatives for excluded users.
Anticipatory Strategy & Experimentation
Bounded pilots, measurable learning, horizon scanning, scenarios and institutional mechanisms for learning from both success and failure.
Validation path
Pilot first, benchmark second
The index will be treated as a measurement instrument, not a marketing questionnaire.
1. Expert review
Target 10–15 experts across AI literacy, responsible AI, higher education, psychometrics, foresight and research methods.
2. Cognitive testing
Target 15–20 students, faculty and leaders to test whether each item is interpreted as intended.
3. Pilot
Target 250–400 respondents from 6–8 diverse higher-education institutions. The pilot will test reliability, item quality, factor structure, missingness and the relationship between self-report and applied judgement scenarios.
4. National fieldwork
Provisional target: at least 2,000 respondents across 40+ HEC-recognized institutions/campuses, subject to access, ethics approval and final sampling calculations.
Scoring guardrails
- People and institutional readiness will be reported as separate primary indices.
- Prototype domains begin with equal weights; weighting can change only after validation.
- Applied judgement scenarios will be reported separately during the pilot.
- Don't-know responses will be reported as an uncertainty signal, not silently converted to neutral.
- No public university league table is planned for Version 1.
- A single combined headline score will not be used unless empirical validation supports it.
Sampling architecture
Recruitment will seek public and private institutions across Pakistan's major regions and broad disciplines. HEC-recognized institution and campus lists will be used as the sampling frame. Any claim of national representativeness will depend on the achieved sample, coverage and weighting rather than on the survey title.
The first release will prioritise national and subgroup patterns. Institution-level dashboards should remain confidential unless sample-quality thresholds are met and the institution explicitly opts into public reporting.
Conceptual foundations
Built to connect AI readiness with anticipatory capability
PAFRI is informed by established international frameworks while developing original item wording and a Pakistan-specific validation process.
UNESCO's AI competency frameworks emphasise human-centred thinking, ethics, technical understanding and responsible use; the OECD/European Commission framework treats AI literacy as knowledge, skills and attitudes for understanding, evaluating and using AI responsibly.
NIST's AI Risk Management Framework contributes governance and risk-management logic. UNESCO's Futures Literacy work contributes the capacity to surface anticipatory assumptions and use multiple imagined futures. The World Economic Forum's Future of Jobs 2025 also identifies AI and big data alongside resilience, flexibility, creative thinking and lifelong learning among rising skills.
Research integrity
What PAFRI will not claim prematurely
- The current protocol is not a validated national index.
- Existing Pakistani AI-readiness research will be cited rather than ignored.
- Self-reported AI use will not be presented as demonstrated competence.
- Small score differences will not be converted into simplistic university rankings.
- National claims will be qualified if the achieved sample remains convenience-based.
Related work: PAUSE-AI Responsible AI Readiness Framework and Futures Literacy in Pakistan.
Contribute to the benchmark
Universities, researchers and experts can contribute to expert review, pilot validation or national fieldwork once the research protocol is approved.