Sri Lanka's university leaders gathered in Colombo on Friday to argue that graduate employability can no longer be measured by placement rates alone. The closed-door roundtable, convened by ETEducation at the Hilton Colombo, brought together vice chancellors, deans and directors general to examine what it means to make graduates ready for an economy reshaped by AI - not merely employed.
The session, themed "Human-Centred AI, Education & Employability: Shaping Sri Lanka's Future Workforce," deliberately moved past the access-versus-enrolment framing that has dominated national education policy for a decade. ETEducation's opening position was direct: access was the last decade's problem, readiness is this decade's, and readiness is now a human question as much as a curricular one. As AI absorbs the technical and procedural work graduates once competed on, the durable differentiator for Sri Lankan talent is shifting toward judgment, ethics, behavioural insight and adaptability.
Participants included Dr DMA Kulasooriya, Director General of the National Institute of Business Management (NIBM); Dr Manjula Wickramasinghe, Dean of the Faculty of Technology at Wayamba University; Dr Sajeewanie Liyana-arachchi, Vice Chancellor of Horizon Campus; Prof Samantha Rajapaksha, Head of the Department of Information Technology at SLIIT; Prof Rahula Attalage, Pro Vice-Chancellor (Academic) at SLIIT; and Prof N Rajeshwaran, Dean of the Faculty of Commerce and Management at Eastern University.
The labour market is ahead of the curriculum
ETEducation opened with the labour-market backdrop the room was asked to respond to. Sri Lanka's labour force stands at roughly 8.55 million, with participation climbing toward 49.4% and unemployment falling for three consecutive years, from 4.7% in 2023 to 3.9% in 2025. Nearly 77% of the employed population works outside agriculture.
The more provocative data point concerned global positioning. Sri Lanka now carries the highest AI exposure of any country in South Asia alongside Bhutan, and AI-skilled roles command close to a 30% wage premium - among the highest in the region. Employers are hiring for AI, data science, cloud, cybersecurity and digital marketing. But what they say is missing from graduate interviews is rarely technical. It is critical thinking, communication, adaptability, leadership and sound judgment. None of those six capabilities show up on a syllabus the way a programming language does.
Placement rates no longer capture institutional performance
The room's institutional data made the point concretely. Prof Rajeshwaran reported placement outcomes among the strongest in the system - more than 80% of graduates employed within six months and effectively full employment within a year, with accounting graduates clearing 99%. Final-semester internship exposure was already converting into permanent offers before students formally graduated.
But the discussion pushed past the headline numbers. Leaders agreed that placement percentage alone no longer captures institutional performance. The more meaningful metric is whether employment is appropriate: aligned to a graduate's qualifications, expectations, capability and long-term career trajectory, not simply any job secured within a reporting window. Employability only became a formally mandated institutional responsibility over the past decade to fifteen years, and approaches to tracking it still vary widely across universities, relying on uneven combinations of employer feedback, accreditation cycles and graduate surveys.
Industry engagement exists - conversion does not
There was broad agreement that Sri Lankan universities are not short on industry exposure. Internships, live client projects, industry-supervised assignments and guest faculty arrangements are already embedded across programmes, including 2+1 and 2+2 models that interleave full-time study with industrial training. Dr Kulasooriya pointed to practices such as third-year internships that shift lecture schedules to weekends, confidential industry-supervisor evaluations, and assignments that require students to interview real employers rather than submit recycled project work.
The gap, participants agreed, is not engagement but conversion - the absence of a uniform, evidence-based system for turning industry feedback into curriculum change and then measuring whether that change worked. With formal curriculum revisions typically occurring on five-year cycles, several leaders flagged the latency between an emerging industry requirement and a graduating cohort equipped to meet it as a structural weakness the sector still needs to solve.
Micro-credentials are useful but not yet trusted at scale
Prof Rajapaksha and Prof Wickramasinghe spoke to a theme that ran through the session: micro-credentials are proving useful precisely because they move faster than formal curriculum cycles, particularly for AI-related capabilities. Graduates who completed additional short courses during the pandemic years reportedly saw stronger employer interest, and IT students are already gravitating toward stackable, short-form learning.
Yet degrees remain the dominant aspiration in Sri Lanka because social recognition and perceived value are still concentrated in full qualifications. Without common standards and formal quality assurance, several leaders cautioned, micro-credentials risk becoming inconsistent or purely revenue-driven rather than a credible complement to degree pathways.
The discussion connected directly to ongoing work in AI for Education, where short-form training is increasingly used to close skills gaps faster than traditional degree cycles allow.
Human-centred capabilities as non-negotiable curriculum
Dr Liyana-arachchi and Prof Attalage pressed the session's central thesis furthest: AI's rapid absorption of technical and procedural work is making sustainability, ethical communication, behavioural understanding, critical thinking and problem-solving explicit learning outcomes rather than assumed by-products of technical training. With the shelf-life of "21st-century skills" now arguably shorter than a decade, the room broadly endorsed curriculum design that incorporates AI quickly while deliberately protecting the human judgment that technology cannot replicate.
Prof Attalage said the shift requires educators to rethink what they assess. "We are no longer training students to compete with machines on speed or recall," he said. "We are training them to do what machines cannot - exercise judgment, understand people, and take responsibility for outcomes."
Entrepreneurship as the other half of readiness
The conversation closed on a theme with direct national-development stakes: employability, participants agreed, should encompass entrepreneurship, remote work and technology transfer, not only conventional salaried employment. University Business Linkage initiatives have evolved into Knowledge and Technology Transfer Centres aimed at converting research into products, spinouts and jobs, yet enrolment in entrepreneurship-focused degree tracks remains low. Cultural preference for secure government or white-collar employment continues to weigh against self-employment.
Leaders noted that government support - grants, incubators, interest-free education loans - exists but remains under-accessed; one recent national programme had disbursed only a fraction of the capital available to it. A three-to-five-year pathway of industry experience followed by a transition into entrepreneurship was floated as a realistic bridge, with stronger job creation positioned as a direct lever against brain drain, without requiring graduates to forgo international careers.
The session drew a comparative reference to India, where industry internships are increasingly mandatory from the first year of four- and five-year degree programmes, and intense competition for limited public university seats has accelerated both micro-credentialing and entrepreneurship. Several leaders suggested Sri Lanka is only beginning to replicate that dynamic.
For educators looking to build these capabilities into their own teaching, structured training such as the AI Learning Path for Teachers offers a practical starting point for integrating AI literacy into course design.
Why this matters for education professionals
For university leaders, faculty and curriculum designers, the roundtable's core message is that placement statistics are no longer a sufficient measure of institutional success. The metrics that will matter next are whether graduates secure work aligned to their qualifications, whether industry feedback loops feed into curriculum change faster than five-year review cycles, and whether human-centred capabilities are taught explicitly rather than assumed as by-products of technical training.
ETEducation confirmed it will develop a follow-up article from the roundtable and continue facilitating industry-university dialogue in Sri Lanka. The session marked the organisation's first on-ground leadership engagement in the country.
Your membership also unlocks:
