THE FUTURE OF TECH-DRIVEN ADMISSIONS TESTING AI-driven assessment has the potential to improve fairness, consistency and accessibility at scale – but only when institutions understand and trust how those systems operate. This was the message from Duolingo, which offers English proficiency assessment designed for international student admissions. Elie Bechara, senior strategic engagement manager for Europe, along with his UK market counterpart Tamsin Thomas, outlined four key points: consistency at scale; a greater focus on human capability; human expertise; and how trust extends beyond score accuracy. “One of the things that we’re really aware of is the ability of AI tests to bring consistency to testing on a phenomenal scale”, noted Bechara. “Not only is it more reliable than humans, but it can also be delivered at a scale never possible before – we’re testing over one million people a year.” Talking about trust, the team highlighted the point that universities need confidence not only in outcomes, but also in governance, explainability, security and the ability to defend decisions publicly. “The more you take humans out of the process, the more you’ve got to build trust and confidence across the stakeholder space.” The session also considered how admissions teams can make principled decisions in an increasingly crowded market, noting that “admissions decisions don’t happen in a vacuum, but in a commercially pressurised environment”. A further finding was that there is strong demand for shorter, more career-oriented courses. Enrique Peñalba, director of operations at IE Talent & Careers, outlined how IE Business School is using AI to scale career services through its AMBA award- winning job-matching tool. The tool uses data from the school’s careers portal, alongside student profiles, CVs and job postings to send hyper- personalised recommendations directly to FROM JOB SEARCH TO JOB MATCH
Dane Anderton (far left), director of MBA programmes at the University of Liverpool, chaired a session on designing specialised programmes
BREAKOUT SESSIONS
FORGING THE FUTURE- READY INSTITUTION In a presentation on academia’s evolving landscape, Martin Bean, founder and CEO of education and leadership consultancy The Bean Centre, outlined six macro trends redefining higher education and global labour markets. These include: • Skills-based hiring over static degrees: employers increasingly prioritise verified, granular skills data over traditional academic credentials to measure job readiness. • Lifelong learning as an economic engine: continuous, modular and stackable capability development is replacing episodic education to drive productivity. • Trust and portability in credentials: as micro-credentials proliferate, verifiable digital infrastructure is essential. • Generative AI as a daily co-worker: this technology is now fundamental infrastructure, creating rapid skill churn and driving massive wage premiums for AI-literate talent. • Enduring human capabilities as the new premium: while routine tasks are automated, judgement, ethical reasoning, adaptability and empathy remain hard to replace. • Relevance linked to revenue: declining demographic cohorts and public funding require universities to diversify revenue through employer partnerships, stackable micro-credentials and lifelong learning. The CEO then turned his attention to strategic imperatives for leadership, asserting that higher education leaders must “urgently redesign learning architectures by shifting focus from knowledge delivery to certifying verifiable capability”. They must also overhaul assessment frameworks “to measure human judgement rather than automated recall”
and integrate AI into “student support, mental health triage and personalised learning pathways”.
ADDRESSING THE AI INFRASTRUCTURE LAG
The constraint in AI today is rarely the horsepower of the model; it is the quality of the terrain we expect it to navigate, explained Sean Traigle, executive vice- president at BoodleBox. Companies are dropping 21st-century supercomputers into 20th-century workflows; however, until organisations stop asking for “faster horses” and start building “better roads” – ie upgrading data architecture, governance and institutional workflows – AI will stay stuck in low gear. Along with colleagues Zach Kinzler and Meredith Lancaster, Traigle explored the core paradigm shift taking place in terms of the traditional workflow, where human effort peaks during the middle phases of problem- solving. The key takeaway: generative AI flattens the execution curve, or the create phase, and consequently, human value shifts from output production to problem definition and quality discernment. The presentation argued that higher education traditionally focuses heavily on teaching students design & create skills. Because AI makes generation almost free, continuing to grade or train primarily on raw execution creates a mismatch with workforce requirements. Institutions must therefore re-orient curriculum around “define” (strategic framing) and “validate” (critical evaluation). The goal is not replacing human effort, but “reallocating cognitive energy toward high-level strategy and verification”.
24 Ambition • ISSUE 4 • 2026
Made with FlippingBook - Share PDF online