by Dr Ahmad Wiraputra Selamat
Not long ago, someone who wanted to learn world history, mathematics, digital marketing or mobile application development had several familiar options: enrol in a university programme, attend a professional course, watch tutorials on YouTube or register for a Massive Open Online Course (MOOC).
Today, that same learner can simply open a generative artificial intelligence (AI) and ask: “Teach me the fundamentals of world history”, “Create a four-week learning plan for Flutter”, “Explain this concept in simpler language” or “Review my work and tell me how to improve it”.
Within seconds, AI can generate explanations, examples, exercises, quizzes and even personalised feedback. For working adults, students and members of the public, this can be faster and is often cheaper than enrolling in a structured online course.
This development raises an uncomfortable but necessary question for universities and training providers: if AI can provide learning on demand, are micro-credentials and MOOCs still relevant?
AI has not eliminated the need for structured learning. It has, however, exposed the weaknesses of micro-credentials that offer little more than recorded videos, downloadable notes, simple quizzes and an automatically generated certificate.
The traditional online course was developed in a period when quality learning content was relatively difficult to produce and distribute. Universities and online learning platforms added value by organising expert knowledge and making it accessible to large numbers of learners.
Generative AI has disrupted this model. Information is no longer the scarce resource it once was. An AI assistant can explain the same concept at beginner, intermediate or advanced levels. It can translate content, generate examples for a specific profession and respond to follow-up questions without requiring the learner to wait for an instructor.
This does not mean that everything generated by AI is accurate, unbiased or educationally appropriate. UNESCO has warned that the rapid development of generative AI is outpacing regulation and institutional preparedness, particularly in areas such as privacy, ethical use and pedagogical validation. Nevertheless, its convenience is difficult to ignore.
If the main value of a micro-credential is access to information, it will struggle to compete with AI. Universities should therefore stop treating content as the final product.
The real product must be a learning experience through which competence can be developed, demonstrated, assessed, verified and recognised.
AI can support learning, but can it establish trust?
A learner may use AI to study programming, accounting, teaching or project management. However, an AI conversation does not automatically provide reliable evidence that the learner can apply that knowledge in a real situation.
This is where well-designed micro-credentials can remain valuable.
A credible micro-credential should provide clear and verifiable information about the specific competencies achieved, the level at which they were demonstrated, who assessed them, and the evidence used. It should also confirm the learner’s identity, show whether the credential is recognised by employers or educational institutions, and indicate whether it can contribute credit towards a larger qualification. In other words, AI can help someone learn, but a trusted institution can verify what that person has learned.
This distinction matters because the need for continuous learning is not disappearing. According to the World Economic Forum’s Future of Jobs Report 2025, employers expect 39 per cent of workers’ core skills to change by 2030. Workers will therefore need faster and more flexible opportunities to update their capabilities.
The paradox is that AI is simultaneously accelerating changes in workplace skills and providing a new way to learn those skills. Micro-credentials can bridge these two developments, but only if they offer more than a certificate of attendance.
The certificate trap
The greatest threat to micro-credentials may not be AI itself. It may be the excessive production of low-value certificates. A certificate does not automatically constitute a meaningful credential. If learners can complete a course by playing several videos, clicking through presentation slides, and answering predictable multiple-choice questions, what does the resulting certificate actually prove?
It may prove participation. It may prove that the learner was able to navigate the platform. It does not necessarily prove professional capability.
This problem existed before generative AI, but AI has made it more visible. Assessments that focus on recalling information or producing generic written answers are now easily completed with AI assistance. Consequently, employers may become even more sceptical of certificates without credible evidence.
The OECD has similarly cautioned that evidence about the employment and social impact of micro-credentials remains limited. Available findings suggest that their value varies significantly according to the programme, provider, field and learner. Micro-credentials are therefore not valuable simply because they are short, digital or labelled as “industry-relevant.”
How universities should respond
Universities should not attempt to compete with AI by producing more content. They should compete through educational design, expert guidance, authentic assessment, quality assurance and institutional trust.
Begin with demonstrable capability
Every micro-credential should be built around a clear question: what should a learner be able to do after completing it? For example, “understanding mobile application development” is too broad. “Developing and testing a functional mobile application that retrieves data through an API” is more specific and assessable. Learning outcomes should describe observable capability rather than mere exposure to content.
Integrate AI instead of pretending it does not exist
Prohibiting AI from all micro-credential activities is unlikely to reflect modern learning or workplace practices. Many participants will eventually use AI in their jobs, so they should learn how to use it responsibly.
A micro-credential can require learners to disclose how AI was used, verify AI-generated information, identify errors, improve weak outputs and justify their final decisions. The purpose is not merely to test whether learners can generate an answer. It is to determine whether they possess the judgement required to evaluate and apply that answer.
Replace easily automated tests with authentic assessment
Assessment should involve meaningful evidence such as a working prototype, workplace project, teaching portfolio, data analysis, live presentation, practical demonstration or oral defence.
Where appropriate, learners could submit both the final product and evidence of the process used to create it. Short interviews or demonstrations can help assess whether the learner genuinely understands the submitted work.
This makes assessment more demanding, but it also makes the credential more credible.
Make credentials stackable and portable
A learner should know whether a micro-credential can contribute towards another course, professional certification or formal academic qualification. At a UNESCO conference on micro-credentials in June 2025, UNESCO Assistant Director-General for Education Stefania Giannini proposed a “three T strategy” which are transparency, trust and transferability, to strengthen the role of micro-credentials in education systems. Without these elements, learners may accumulate certificates that cannot be used beyond the platform or institution that issued them. UNESCO’s report explains that transparency concerns clarity about learning outcomes and assessment, trust depends on reliable quality assurance, and transferability enables credentials to support further study, mobility and employment.
Malaysia, for example, has established an important policy foundation for micro-credentials. The Malaysian Qualifications Agency introduced its Guidelines to Good Practices for Micro-credentials in 2020, followed by the Guidelines to Good Practices: Quality Verification of Stand-Alone Micro-credentials in 2023. More recently, MQA issued its Guidelines on the Academic Bank of Credit in 2026 to facilitate the accumulation, storage, transfer and utilisation of academic credits. Such infrastructure is important because the long-term value of micro-credentials depends on whether learning can be accumulated, transferred and recognised.
Publish evidence of outcomes
Universities should report more than enrolment and completion numbers. They should examine whether participants used the acquired skills, obtained academic credit, progressed to further study, received new responsibilities or improved their employment opportunities.
Not every micro-credential must lead directly to a new job. Some may support personal development, community learning or continuing professional education. Nevertheless, providers should state the intended value clearly and collect evidence showing whether that value is being achieved.
Industry cannot remain a passive observer
Industries frequently ask universities to produce more job-ready graduates and offer courses aligned with current skills. However, employers must also participate in defining, delivering and recognising micro-credentials.
Industry partners can contribute authentic problems, datasets, professional standards, mentors and assessors. They can identify the level of performance expected in the workplace and help universities keep rapidly changing content current.
More importantly, employers must communicate whether they genuinely recognise particular micro-credentials during recruitment and promotion. A claim of “industry relevance” means little if employers continue to consider only conventional degrees and years of experience.
Employers should also look beyond the number of certificates listed on a resume. A better question is what can the applicant demonstrate, and what verified evidence is attached to the credential?
Access and inequality must not be ignored
AI may appear to democratise education, but access to technology does not automatically create equal learning opportunities.
Experienced learners may know what questions to ask AI, how to verify its answers and how to organise their learning. Beginners may not. Some learners require structure, motivation, feedback, peer interaction and access to a knowledgeable instructor. Language, connectivity, disability, cost and digital literacy can also affect who benefits.
Well-designed micro-credentials can provide this support while remaining more flexible than conventional degree programmes. Poorly designed ones may simply create another market in which learners pay for certificates without receiving meaningful educational guidance.
Therefore, the social purpose of micro-credentials should remain central. They should expand access to credible learning, not merely expand the number of products universities can sell.
From content delivery to trusted evidence
The future is not a competition between AI and micro-credentials. The more useful model is an AI-enhanced micro-credential.
AI can provide personalised explanations, practice activities, simulations and immediate formative feedback. Educators can provide context, challenge assumptions, guide ethical practice and design authentic learning experiences. Institutions and industries can jointly verify and recognise the resulting competencies.
The division of responsibility is important. AI can assist the learning process, but accountable institutions must safeguard the credibility of the credential. AI will not kill micro-credentials. It will kill micro-credentials that offer nothing beyond information and a certificate.
Those that provide guided practice, authentic assessment, portable credit, industry recognition and trustworthy evidence of competence will continue to matter. The question is no longer whether learners can access knowledge. The question is whether universities and industries can credibly demonstrate what learners are able to do with it.
Dr Ahmad Wiraputra Selamat is a Software Engineering lecturer and Coordinator of e-Learning at the Academic Development Centre, Universiti Pendidikan Sultan Idris (UPSI), Malaysia. His interests include digital education, micro-credentials, MOOCs and AI-assisted teaching and learning.
Leave a Reply