Adoption of AI in education is accelerating. Massive potential but hurdles remain.

The HolonIQ Global Executive Panel reports strong early adoption of artificial intelligence, emerging capability building, but still many barriers.

Neural Networks

377 respondents from the Global Executive Panel report growing adoption of AI and emerging capability building.

Education is perhaps one of the most obvious areas for the application of artificial intelligence, with the potential to improve access, dramatically reduce cost and accelerate learning outcomes. As in many other fields, AI in education is focused on performing cognitive functions such as perceiving, reasoning, learning, and problem-solving with application to vast the challenges and opportunities of continuously educating the 7.7 billion people on the planet.

Overall, the Global Executive Panel sees significant potential for AI in education, with the greatest impact expected in testing and assessment. Post-secondary and language learning were also considered sectors which will be impacted, followed by K12. There were a wider range of views around AI’s potential for Pre K.

However, it is still very early days for adoption. Only 1 in 10 of the Global Executive Panel’s organizations have deployed AI, nearly 1 in 5 have completed a pilot and 1 in 3 have no action planned. Despite early days, the results suggest that AI is already generating meaningful value. 40% of respondents reported significant value and 40% reported moderate value has been gained from AI’s use in learning processes.

A lack of clear AI strategy was cited as the biggest barrier for adoption, with a lack of talent, data and leadership commitment all noted as challenges in moving from aspiration to adoption. Just under 1 in 3 organizations were not currently building AI capability and of those who are, in-house skills development, partnering and licensing are being used as alternatives for capacity building.

Of the eleven capabilities for creating value at scale through AI, one quarter of respondents noted that their organization had none of these capabilities in place.

On the whole, responses indicate much potential for AI in education, a strong cadre of early of adopters achieving significant value in predominately learning processes, alongside not to be under-estimated challenges to enabling and mobilising AI for the education sector as a whole.

Screenshot 2019-04-01 21.39.16

About the Global Executive Panel

HolonIQ’s online AI in Education survey was in the field from 18 to 29 March 2019 and garnered responses from 377 executives from over 25 countries, and representing the full range of education institutions, education technology and services companies and enablers such as government, investors, accelerators and incubators. The HolonIQ Global Executive Panel is an invitation-only expert group of education and technology executives from around the world who are the leaders, engines, entrepreneurs and intellectual power shaping the future of education.

Request an invitation to join the panel here.

Screenshot 2019-04-01 22.10.45

Greatest AI impact for testing and assessment

The results for potential impact by sector and technology show the median score from the Executive Panel with the bottom and top quartile, giving a little more insight into the distribution of views.

Testing was seen by most as having a very high AI potential and impact. Language Learning and Post Secondary, from Technical and Vocational Training through Higher Education and Corporate Training, were all considered high potential and high impact. K12 followed close behind however the Executive Panel shared a wide range of views for lower potential and impact in the Pre K space.

Screenshot 2019-04-01 22.10.45
Screenshot 2019-04-01 21.52.15

High value impact for algorithms, less use of vision based AI

HolonIQ’s AI technology Applications Framework identifies five key technologies driving uses of AI in education – Vision, Voice, NLP, Algorithms and Hardware.

When asked to consider which of these technologies were considered the leading source of impact, the Executive Panel identified Algorithms followed by Natural Language Processing/Linguistics. Voice-based AI applications followed with a broader spread on higher impact with a median impact score of 7 shared with hardware and vision. Hardware has a tighter set of views than Vision based AI which is more spread towards lower impact/potential than peer technologies.

AI adoption Plans and Actions

Adoption of AI and plans for adoption differ across the various organization types represented by the Executive Panel. In general however, 1 in 10 organizations have invested in and deployed artificial intelligence in some part of their operations. While almost 20% are conducting a pilot, around 1 in 3 have AI on the radar but currently have no specific plans for adoption. Most of the remainder have AI in short or medium term planning and 5% responded that they have no interested in adopting AI at all.

Where is AI being used?

The panel shared a more granular perspective on where the adoption of AI has occurred inside their organizations. The results generally follow perceptions on impact and potential except in the case of Vision based AI which was considered having the least impact and potential yet leads Voice and Hardware in terms of adoption. Algorithms and NLP have seen the greatest levels of adoption. Perhaps unsurprisingly, EdTech companies are the clear leader in AI adoption. Services and Enablers have 5-10% lower adoption rates with education institutions not far behind.

Why is AI being used?

Reinforcing the view that use of AI is still in the early adoption phase, two out of three organizations are adopting AI with the aim to disrupt the market, seeing the potential of AI to change the game and desiring to be one of those leaders. However, as the benefits of AI start to shape into competitive advantage we are seeing more organizations using AI for cost savings and to drive efficiency.

Similarly, as the value of AI technology starts to penetrate the user/customer/learner experience, responding to demand from clients will become a more important driver, with an average of one in three citing customer demand as a reason for AI adoption, up to 50% for those from EdTech organizations.

Improving learner experiences and organizational agility are also high on the list as drivers for the adoption of AI technologies.

Where is AI creating most value?

While the adoption of AI in education is still in its early days, the results suggest that where it is deployed, value is being generated.

When the Executive Panel was asked about the value captured from AI deployment, 40% reported significant value and 40% reported moderate value was created through the use of AI in learning processes which was the area of highest value generation.

Student/Customer support followed closely behind along with Assessment and Feedback, then Student/Customer Acquisition. All six of the areas identified achieved relatively high levels of value creation, including Identity and Security, which was ranked lowest in value creation.

Progress on enablers of AI at scale

To take advantage of AI’s enormous potential, most organizations have a long way to go in developing the core practices that enable them to realize the potential value of AI at scale.

Less than 20% of respondents say their organizations have mapped out where all potential AI opportunities lie and only one quarter indicated that their organization has a clear strategy in place for sourcing the data that enables artificial intelligence.

Finally, nearly a quarter of respondents say their companies have not developed any of the eleven AI enabling practices, indicating that it may take some time for the industry as a whole to gain maximum value from the use of artificial intelligence in their operations.

What are the barriers to adoption of AI in education?

Almost half of the respondents reported that the biggest barrier to adoption of AI in education is the lack of a clear AI strategy, differing by type with institutions (schools, universities, colleges etc.) at 50% and EdTech organizations at 25%. Both groups cited a lack of AI talent as a major barrier (45% of respondents) representing the second largest barrier overall. Twenty percent of institutions and one third of EdTech organisations identified the lack of data as a major barrier, perhaps indicating a fragmented systems environment or data policy issues. The vision for and use of Artificial Intelligence crosses almost all functional boundaries and can come with significant costs, so it’s not surprising to see leadership and technical infrastructure identified as barriers to adoption.

How are organizations building AI capability?

One third of organizations represented on the Executive Panel are not currently developing AI capabilities, with the EdTech an outlier at only 15%. This result is perhaps not surprising given responses about lack of AI strategy and leadership, however is cause for concern in a sector that

Of those who are building AI capabilities, 1 in 3 is developing this talent in-house or partnering to develop capacity (50% for EdTech companies, with Services and Enablers behind institutional respondents).

30% of institutions are buying or licensing AI capability, perhaps signalling their use of EdTech with in-build AI functionality and EdTech more likely to be recruiting on the open market for AI skills than other organizational types.

Globally across all industries the skills gap for advanced technology talent such as artificial intelligence is well documented. While the education sector faces the same problems as other industries, it is the sector which is perhaps best positioned to upskill the next generation of tech talent.

Looking ahead...

These insights about Artificial Intelligence in Education are thanks to the HolonIQ Global Executive Panel, made up of over one thousand professionals, leaders, academics and entrepreneurs working in or with the global education sector and who represent the intellectual power shaping the future of education.

Become a Client

HolonIQ is the world’s smartest source of global education intelligence to power decisions that matter. Learn more about how we can help you achieve your mission-critical priorities.

space

Related Research

2019 Robotics in Education Report

29 August 2019. 

HolonIQ’s 2019 Robotics in Education Report includes Market Maps and Segments, Growth and Key Players.

Robotics in Education. An AI-powered future.

8 August 2019. 

The Robotics market in education is forecast to grow from $1.3B to $3.1B to 2025, at which point it will be displaced by Artificial Intelligence.

Education-as-Usual. Likely but not preferred.

6 June 2019. 

Of the 5 Scenarios for Education in 2030, Education-as-Usual is least preferred but most likely according to 5000+ educators, entrepreneurs, leaders and investors from 30+ countries who have contributed to the Education in 2030 Project.

2019 Artificial Intelligence & Global Education Report

24 May 2019. 

HolonIQ’s Annual Report on the State of Artificial Intelligence in Global Education