Increased computational power and massive amounts of data have made Artificial Intelligence (AI) and Machine Learning (ML) a key driver of growth for many industries. In Education, AI has seen rapid progress in recent years as it aims to develop new teaching and learning solutions to improve education quality, increase efficiency, and promote educational equity.
HolonIQ's ongoing research into AI in Education explores the potential impact of AI technologies across education sectors globally, including expectations and approaches to adopting AI, application and value creation, organizational barriers, and capability development. This research update captures education industry attitudes and shifts as institutions and organizations consider the role of AI in digital adoption and transformation progress.
AI in education is expected to benefit learners, educators and institutions in multiple ways, including automation of administrative tasks and processes and personalized learning at scale, enabling educators to focus on higher value activities such as student mentoring and support.
Expectations for AI impact: technologies and markets
HolonIQ’s AI technology Applications Framework identifies four key technologies driving uses of AI in education: Vision, Voice, Language and Analytics.
At this point in time, Analytics and Language are perceived as having most potential impact on the education industry. Both applications play an integral part, for example, in scaling and integrating intelligent Adaptive Learning solutions across the education sector. Intelligent Adaptive Learning is a disruptive technology backed by AI that aims to personalize teaching based on students' learning preferences, knowledge state and progress.
"We will be using machine learning as part of what we do in the next two years which should vastly improve customer experience around our core functionality."
Figure 1. Assessing potential impact of Voice, Vision, ML & NLP technologies on education
While it is still in the early stage, Voice and Vision are also important sub-applications for Intelligent Adaptive learning systems. For instance, computer vision has been increasingly used to gather information such as mouse movements, eye tracking and sentiment analysis when teaching complex subjects, delivering in-depth insights on the student performance. Since the COVID-19 pandemic, these solutions have also been introduced through technologies such as remote proctoring.
Figure 2. Expected impact of AI technologies on different education markets.
Looking at education markets more broadly, AI is expected to have most impact on Testing and Assessment, followed by Language Learning, Corporate Training/ Upskilling and Higher Education.
Access the Full Briefing
Customers can access the full deep dive on platform
Organizations are picking up the pace in their planning and adoption of Artificial Intelligence. 25% report successful investment and deployment of AI in 2022, compared to 14% in 2019. AI is gradually shifting from consideration to action; 44% of respondents have included AI in their short to medium term plans, and although pilots are still happening, these have decreased since 2019. Likewise, those who consider AI to be 'on the radar' but are not planning action have decreased from 24% in 2019 to 14% in 2022.
Figure 3. Organizational planning for AI: 2019 vs 2022
"Like many institutions we are in the planning stage. There are not many solutions out there apart from reporting and analysis of data. Solutions with learning analytics that are predictive are limited."
Looking at specific technologies, education organizations appear to have made most progress embedding Machine Learning into their education operations (38%), compared to 29% in NLP and 24% in Vision.
Reports vary for Vision and Voice applications, with more than a third of respondents having no current plans in place for AI application. These technologies are still relatively new for many organizations, and many tools and applications are still in stages of development or piloting use in different contexts.
AI adoption rationale and expected value creation
The most common reason for adopting AI among our respondents was to improve customer outcomes. 75% of our sample noted this as a priority, compared with 45% who cited cost reduction, and 43% looking to disrupt the market.
The accelerated shift to online learning over the last few years has placed the customer (student) experience in sharp focus for education institutions. Some AI-driven solutions aim to improve student satisfaction and outcomes with personalized learning and support, for example, as well as enabling teachers to design and deliver learning in more efficient and innovative ways.
Overall, there is optimism among both institutions and digital providers for the potential of AI to drive innovation, improve agility and disrupt established ways of operating in education markets and institutions.
Figure 4. Reasons for adopting AI in education
"AI has the potential to transform and optimise the way we work in higher education and our students' learning experiences. It's still in its very early stages at my university, but there is planning in place and leaders who understand the value."
As stakeholders consider the emerging value AI may create, there are high expectations for areas of education where assessment and feedback is central to operations. Around a third of our sample expect significant impact here, and a further 31% expect moderate impact on value.
Figure 5. Where is AI expected to create value?
AI technologies are also expected to bring improved value to Learning Processes and Customer Support, pointing to different parts of the learner experience and lifecycle. AI in Customer Support may bring improvements in marketing, recruitment and enrolment, as well as student support throughout their learning experience. Learning Processes are likely to deploy AI across learning design and learner experience, bringing potential efficiencies but also new ways of designing and delivering learning in different educational contexts.
"AI affects every aspect of student journeys as well as teachers day-to-day lives, scaled to millions of students."
Enablers and barriers to AI adoption
To take advantage of AI’s potential, many organizations have started to include core AI practices and tools that enable them to realize the value of AI at scale. 43% report that senior leaders have shown significant commitment to adopt AI initiatives, and a further 47% note that their organizations are using data effectively to support goals of AI work.
Figure 6. AI practices to enable adoption
Whilst some of these enabling practices are similar to those reported in 2019, others have shifted. There has been an increase, for example, in organizations reporting that they have mapped where all potential AI opportunities lie (40% in 2022, compared to 27% in 2019). On the other hand, fewer are likely to feel that they have access to the talent and skill sets to support AI work now (28% in 2022, vs. 37% in 2019); this may be due to several factors, include global labor shortages in digital skills, as well as increasing complexity and capabilities of the AI tools themselves, which demand new and different talent from within and outside the organisation.
Figure 7. Barriers to adoption
As companies invest and transition towards an AI-driven education ecosystem, they are finding it increasingly hard to recruit the right talent to work on these key technology challenges. 54% reported this as the leading barrier limiting AI adoption plans, compared to 44% in 2019. The second biggest barrier (50% of respondents) was cited as 'under-resourcing for AI', which has risen sharply as an issue since 2019 and is likely compounding the issue of recruiting skilled talent.
"One of the biggest challenges for adopting AI insights in operations is the lack of data literacy across the organization."
Another common and rising theme is a lack of clear organizational strategy in AI, noted by 47% of respondents as a barrier restricting AI adoption. Uncertain/ low expectations on investment returns (25%) and significant infrastructure and other functional costs (31%) are some of the additional challenges limiting adoption of AI in education.
Access the Full Briefing
Customers can access the full deep dive on platform
Talent issues become even more pressing when organizations are focussed on developing capabilities in-house, which is currently the most commonly-reported method of developing AI capabilities (40%). At the same time, there have been reductions in those reporting partnerships with others (e.g. academic institutions, foundations) to find talent, and a large drop in organizations reporting that they are buying or licensing capabilities from large technology companies.
Figure 8. Building AI capabilities
"One of the biggest challenge for adopting AI insights in operations is the lack of data literacy across the organization."
Ready to get started?
Get a personalized walkthrough of HolonIQ's Global Intelligence Platform.