COVID-19 posed plenty of challenges for folks around the world, but parents of school-aged children had a particularly hard time. As millions of students were forced to attend virtual school, many parents were too busy with their own WFH woes to keep constant tabs on what their kids were doing. Many schools, likewise, were ill-prepared for the realities of digital learning, and keeping kids engaged in virtual lessons was difficult.

It came as no surprise, then, when declines in learning relative to previous years were observed in students. In June 2021, market research firm Ipsos surveyed US parents with school-aged kids on the future of education. Prior to the pandemic, 81% said their kids’ learning was adequate. But that faith in the system deteriorated as the pandemic dragged on—just 59% felt confident that their children’s schooling was meeting standards in 2021. Even if the pandemic comes to an end, 55% of parents expressed concern that their kids might not be able to catch up.

UK-based economic research firm Institute of Fiscal Studies surveyed parents in England between April and May of 2021 and found that around one in four thought it would take more than a full school year for their kids to catch up to where they would have been without the disruptions. And despite a widespread return to in-class learning, many parents still feel they need to fill in the gaps created by time outside of the classroom.

A report from research institute The Learning Counsel found that US families spent 25% more on curriculum resources and networks in 2020 than they did in 2019, for a cumulative spend of US$35.8 billion.

It’s a growing market opportunity, and ed-tech companies using AI-assisted adaptive learning technology may be well-poised to make the most out of this heightened demand for personalized education.

LA-based Age of Learning—the company behind ABCmouse Early Learning Academy—uses artificial intelligence (also known as machine learning) to identify where children are in terms of their personal trajectory, says chief innovation officer Sunil Gunderia. For US$9.95 per month, the subscription service for two to nines offers programs for reading, math, science and art.

With AI, Age of Learning’s My Math Academy program measures a child’s progress, pulling data from their current knowledge, strengths and weaknesses in order to create customized lesson plans. The personal learning pathways may dive deeper into topics kids are struggling with, or may advance a child to the next learning module early if it identifies they have a strong grasp of the topic. The AI also tracks whether a child has mastered the lesson to determine if they can move forward to the next session. Each is tailored to the individual, and includes pre-recorded sessions, direct instructional content and real-time feedback, says Gunderia.

As more kids head back to school in person, Gunderia says teachers can access the digital education programs for free in their classrooms. The same artificial intelligence that delivers individual personalized learning tools allows educators to track their students’ goals and provides individualized programs for each child or for the entire group.

Madrid-based Lingokids has been using similar machine learning technology on its educational play app since it launched in 2016, according to company CEO Cristobal Viedma.

Lingokids’ machine learning technology reccomends games for kids based on their age, interests and skill level.

For US$14.99 a month, the subscription education service targets children ages two to eight and covers topics like literacy, reading, writing and mathematics.

Lingokids’ AI technology creates “learning pathways” that help kids acquire knowledge​​progressively. Similar to Age of Learning, these customized routes recommend games and activities for a child depending on their interests, age, completion rates and the skills they need to reinforce—as identified by the AI.

The technology collects data that shows Lingokids what components users like the best and what formats they are interacting with the most, allowing it to further refine its offerings.

To ensure the safety of their users’ information, Viedma says Lingokids doesn’t collect pictures or addresses, though it does need certain data points—like a child’s age, for example—to make sure they receive content that is age-appropriate.

While AI requires specific types of information to operate (i.e. did the child answer correctly, are they completing the lesson plan), Viedma says the biggest challenge with the tech is that Lingokids doesn’t hear from children directly.

“Since the algorithm is based on interest and on the child’s progress, the challenge here is that we don’t get to ask the child what exactly they prefer and why,” he says.

The limitation isn’t slowing its growth, however. Earlier this year, Lingokids raised US$40 million, which it will use to expand its international presence and broaden its offerings in science, technology, math and core subjects. Some of the money has been earmarked for new hires to bolster the content, engineering and development teams. Lingokids also plans to expand its emotional intelligence, empathy and critical thinking programs by developing audio content focused on mental health and mindfulness, says Viedma.

Another company that is rapidly expanding globally is Bangalore-based BYJU’s, which has been using AI to offer personalized learning for kids and teens since 2014.

“Personalized learning has a significant impact on our ability to improve outcomes, as children develop at varying paces. Our goal is to provide students of all abilities the instruction they need at a style and pace that suits them best,” says Teri Rousseau, VP of strategy and marketing.

With more than 100 million global users, BYJU’s offers monthly subscriptions for grades one to three (US$46), grades four to 10 (US$400) and older children (between US$460 and US$1,120).

By tracking a user’s progress to determine how well they understand concepts, Rousseau says BYJU’s technology leverages AI to help students learn at home as if a teacher or parent were guiding them step by step.

In BYJU’s Magic Workbooks, for example, the AI scores a child’s worksheet to determine completion and provides kids with real-time feedback from the app’s characters. The characters acknowledge if the student has answered a question correctly, prompting them to try again if not.

Seeing a significant demand for personalized learning, BYJU’s has been on a buying spree. It most recently acquired Austria-based online learning platform GeoGebra, an interactive online math and science app that offers lessons from elementary to university level. While the education platform will continue to operate independently, BYJU’s plans on integrating GeoGebra’s learning tools (such as math tests, apps and ready-made courses) into its own services, and will explore new products and formats as well.

“Until recently, education was very standardized with a one-size-fits-all model,” says Rousseau. “But personalization is no longer the exception; it is becoming the expectation. In the future, there [will be] even deeper levels of personalization. I believe we will be able to provide the right type of response, feedback and tutorials that meet the child where they are at and deliver the help they need in the style of learning that suits them best.”