No funded national programme for AI-era assessment redesign in higher education

HEPI's 2026 survey: 94% of students use generative AI in assessed work, 12% submit AI-generated text directly (up from 3% in 2024), and 22% admit AI-assisted cheating. QAA publishes guidance but is membership-funded in England with no regulatory force; Jisc's National Centre for AI runs a one-year assessment pilot (2025–26); the OfS has no funding stream for redesign. So 140+ financially distressed providers each reinvent assessment alone, mid-redundancy. There is no shared assessment bank, no pooled capacity for scaled vivas or interactive assessment, and no common verification standard for degrees.

Degree credibility is a national export (international students) and the signal underpinning the graduate labour market. Institution-by-institution improvisation under financial distress risks divergent standards and quiet grade-integrity erosion the sector cannot afford.

An OfS/DfE-funded Assessment Transformation Programme: institutional redesign grants, shared toolkits and assessment banks, pooled viva/interactive-assessment capacity, scaling Jisc's pilot and QAA guidance into funded delivery.

// Build together: Counterparty: Jisc/QAA plus volunteer universities pooling assessment banks and viva capacity; OfS/DfE funding is the scaled end-state.

94% of students now use generative AI in assessed work while 140-plus financially distressed providers each improvise alone, with no OfS funding stream or shared standard.