The Potential of Artificial Intelligence and Emerging Technologies for

Digital Accessibility in Childhood Literacies: A Critical Review

of the Literature

Jennifer Milena Bueno-Rocha

, Anuarith da Rosa Joaquim Martins

, Ana Margarida Almeida

and Maria João Antunes

Digital Interaction and Media Research Centre, University of Aveiro, Aveiro, Portugal

Keywords: Emerging Technologies, Artificial Intelligence, Childhood, Literacies, Accessibility, Digital Accessibility.

Abstract: This review examines how AI and emerging technologies can increase digital accessibility in childhood

education, focusing on ethics, communication, tutoring, and health. Findings are presented with consideration

for accessibility and inclusion, and implications for stakeholders are explored. Concerns center around data

protection and children-centered AI development. Opportunities and threats are highlighted based on current

guidelines and frameworks. Despite the potential of AI to bridge social gaps in childhood education, a local

approach that prioritizes contextual needs is crucial.

1 INTRODUCTION

The XXI century is marked by rapid technological

advancements that are transforming almost every

aspect of human life and society. Artificial

Intelligence (AI) is one technology that challenges the

foundations of human intelligence, free will, and

cognition. The increasing sophistication of data

processors has enabled the development of

technologies that can emulate human abilities,

ranging from robots that move like humans to

software that simulates the human mind.

This review explores AI's potential to increase

digital accessibility and literacy development in

childhood education. The authors bring together a

diverse range of disciplinary and cultural

backgrounds, with interests ranging from the impact

of the digital divide on education and health to

eHealth literacy and its influence on health decision-

making. The authors' shared interest in the topic is the

basis for a transdisciplinary dialogue (Shahamiri &

Thabtah, 2020).

The definition of AI adopted in this review is from

UNICEF's policy guidelines (UNICEF & Ministry for

https://orcid.org/0000-0001-6526-8460

https://orcid.org/0000-0001-9759-4930

https://orcid.org/0000-0002-7349-457X

https://orcid.org/0000-0002-7819-4103

foreign affairs of Finland, 2021), which defines AI as

machine-based systems that can predict, recommend,

or decide to influence real or virtual environments.

These systems interact with humans and adapt their

behavior through learning from their context.

According to Wikipedia (Wikipedia Collaborators,

2021), the field of AI is the study of "intelligent

agents" that perceive their environment and take

actions that maximize their chances of achieving their

goals.

This review focuses on AI's impact on formal

education and literacy development, which refers to

acquiring literacy skills inside and outside

educational institutions. Accessibility encompasses

cultural, linguistic, gender, race, and disability-based

and social-based conditions.

This paper discusses the impact of AI on early

childhood education in four sections. The first

provides an overview of research methods and

findings. The second analyzes AI's ethics in

childhood education, communication, language

acquisition, tutoring, and health literacy. The third

examines global trends in policy guidelines. The

fourth explores AI's impact on childhood education,

104

Bueno-Rocha, J., Martins, A., Almeida, A. and Antunes, M.

The Potential of Artiﬁcial Intelligence and Emerging Technologies for Digital Accessibility in Childhood Literacies: A Critical Review of the Literature.

DOI: 10.5220/0011771400003470

In Proceedings of the 15th International Conference on Computer Supported Education (CSEDU 2023) - Volume 2, pages 104-112

ISBN: 978-989-758-641-5; ISSN: 2184-5026

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

health outcomes, and stakeholder groups.

Conclusions provide final insights, while we expose

our challenges and restrictions in Limitations.

2 STATE OF THE ART

Reviewing two kinds of literature ecosystems was

necessary to achieve the goal of this document. First,

the scientific and academic literature, and second, the

regulatory and normative literature produced by

governments, multilateral and supranational

structures (UN, UNICEF, UNESCO, among others),

and expert concept notes from the private sector or

non-governmental organizations (NGOs).

2.1 Methods

We used the critical literature review approach, but

due to limited time and resources, there are some

variations from the PRISMA Statement, particularly

in the double-blind process.

2.1.1 Literature Review

Scientific and Academic Literature

Three scientific databases were selected to cover a

range of disciplines: Scopus, ERIC, and PsycInfo,

along with grey literature on Google Scholar and

Google from 2016 to 2021. Guidelines from each

database were followed to create queries (see Table 1).

Table 1: Elements of the critical literature review approach.

Element Descri

tion

Databases Sco

us, Ps

cInfo, ERIC

Gre

literature Goo

le schoola

Languages English

Perio

Five years (2016-2021)

Document

Original papers, reviews, reports,

uidelines

We formulated a question using the PICO model

to generate relevant keywords and incorporate a

broad range of knowledge fields (Table 2).

What is artificial intelligence and emergent

technologies' potential in digital accessibility on

childhood education and literacies development?

Table 2: Literature review question structure.

P Childhood

I Artificial intelligence technologies

C Traditional interventions for childhood education

Accessibility and Digital Accessibility in education

and literacies development

We used the following keywords to build queries

for databases and search engines based on this

question: Childhood, artificial intelligence,

accessibility, and education (Table 3).

The data-collected documents were analyzed

using the Qatar Computing Research Institute web-

based application Rayyan.ai designed for screening

and analysis purposes in literature reviews (Ouzzani

et al., 2016).

All documents that did not address AI or

emerging technology interventions related to

childhood education or literacy development and

theoretical or position documents were excluded.

Table 3: Queries.

S Query / Queries

PSYCinfo

((Artificial intelligence OR emergent

technologies) AND digital accessibility)

ERIC

"Artificial intelligence" AI digital accessibility

childhood

SCOPUS 1

( ( ( "Artificial Intelligence" OR "AI" ) OR (

"emergent technologies" ) ) AND ( "Digital

accessibility" OR "accessibility"

) AND literacy AND child* ) AND ( LIMIT-

TO ( PUBYEAR , 2022 ) OR LIMIT-TO (

PUBYEAR , 2021 ) OR LIMIT-TO (

PUBYEAR , 2020 ) OR LIMIT-TO (

PUBYEAR , 2019 ) OR LIMIT-TO (

PUBYEAR , 2018 ) OR LIMIT-TO (

PUBYEAR , 2017 ) OR LIMIT-TO (

PUBYEAR , 2016 ) ) AND ( LIMIT-TO (

DOCTYPE , "ar" ) OR LIMIT-TO (

DOCTYPE , "re" ) ) AND ( LIMIT-TO (

LANGUAGE , "English" ) OR LIMIT-TO (

LANGUAGE , "Spanish" ) OR LIMIT-TO (

LANGUAGE , "Portuguese" ) )

SCOPUS 2

TITLE-ABS-KEY((("artificial intelligence" OR

technolog*) AND "health"AND litera* AND

child* AND accessibility)) AND ( LIMIT-TO (

PUBYEAR,2021) OR LIMIT-TO (

PUBYEAR,2020) OR LIMIT-TO (

PUBYEAR,2019) OR LIMIT-TO (

PUBYEAR,2018) OR LIMIT-TO (

PUBYEAR,2017) ) AND ( LIMIT-TO (

DOCTYPE,"ar" ) OR LIMIT-TO (

DOCTYPE,"re" ) OR LIMIT-TO (

DOCTYPE,"ch" ) OR LIMIT-TO (

DOCTYPE,"bk" ) )

Grey Literature

AI and emergent technologies are critical topics for

decision/policymakers, with multiple stakeholder

groups involved in shaping their impact on society.

These stakeholders can be categorized as public

The Potential of Artiﬁcial Intelligence and Emerging Technologies for Digital Accessibility in Childhood Literacies: A Critical Review of

the Literature

105

(including multilateral and regional organizations, as

well as local governments), private (including global

technology conglomerates), and third-sector

(including civil society organizations and academia).

We used Google Search Engine to gather relevant

documents, focusing on keyword combinations with

geographic specifications by continent.

2.2 Findings

2.2.1 State of the Art

In this systematic review, we analyzed 20 full-text

papers on AI applications in childhood education,

following an SLR approach to summarize trends and

notable advances in each topic to get a broad view of

the available literature. The recent research focuses

on four main areas: ethics of AI in education and

childhood, tutoring systems, language translation

systems, and intelligent virtual assistants. Among the

analyzed papers, 32% related to the ethical

implications of AI in education, followed by 32% on

developing tutoring systems and intelligent virtual

assistants. Around 28% of the papers were on

practical applications of AI in human language

processing, and 8% were on AI applications for health

conditions impacting children's education.

2.3 Analysis

We analyzed research and policy documents from

2015 to 2022 to identify trends in the potential of AI

for digital accessibility in childhood literacy

development. The documents reviewed highlight four

main thematic paths.

2.3.1 Ethics Analysis of AI in Childhood

Education

The use of AI in childhood education raises concerns

about the potential and risks of deepening inequality.

Veinot et al. (2018) pointed out that there is a risk of

intervention-generated inequalities, where privileged

individuals and communities may benefit more from

using AI in education than others who are less

privileged. (Veinot et al., 2018). Saltman (2020)

analyzes the development of for-profit AI educational

technologies in the current economic and ideological

structure and its impact on public education

privatization (Saltman, 2020). Park et al. (2021) argue

that the lack of critical thinking development

increases the risks of AI technologies that support

students in their self-directed learning process,

transforming the traditional view of the student as a

passive agent. Educators need to foster critical

thinking development as AI technologies already

support students in their self-directed learning

process, transforming the traditional view of the

student as a passive agent in the teacher-student

relationship (Park et al., 2021). AI research in

childhood education focuses on language acquisition,

speech, hearing, writing, and reading. Natural

Language Processing (NLP) has a broad range of

applications, including text lexical adaptation,

making written texts more accessible to low-literacy

adults, children, and those with cognitive disabilities.

The research in this area can be classified into two

branches: text simplification and text production

(Hartmann & Aluisio, 2021). Deaf children's

vocabulary acquisition through sign language is being

made easier with the development of AI-based

systems like SiLearn (Joy et al., 2019). Additionally,

the production of educational materials for vision-

impaired and blind children is being studied with the

help of AI systems that can scan images and convert

them into 3D-printed braille-tagged versions (See &

Advincula, 2021).

2.3.2 Human Communication: Language

Acquisition, Translation & Processing

AI research in childhood education focuses on

language acquisition, speech, hearing, writing, and

reading. Natural Language Processing (NLP) has a

broad range of applications, including text lexical

adaptation, making written texts more accessible to

low-literacy adults, children, and those with cognitive

disabilities. The research in this area can be classified

into two branches: text simplification and text

production (Hartmann & Aluisio, 2021). Deaf

children's vocabulary acquisition through sign

language is being made easier with the development

of AI-based systems like SiLearn (Joy et al., 2019).

Additionally, the production of educational materials

for vision-impaired and blind children is being

studied with the help of AI systems that can scan

images and convert them into 3D-printed braille-

tagged versions (See & Advincula, 2021).

2.3.3 Tutoring: AI-Based Virtual Intelligent

Assistants Applications

The use of AI systems in teaching and tutoring has

become a significant area of research in childhood

education. While the ideal learning environment

would have a team of professionals to support

students with specific learning needs, such an

environment is rare in most schools worldwide.

CSEDU 2023 - 15th International Conference on Computer Supported Education

106

Therefore, AI-based systems have been studied to

accompany students with special needs or disabilities.

IBM's Watson Tutor (WT) system is a recognized

AI-based platform designed to improve educational

outcomes by promoting student engagement with

content through a dialog strategy in a one-on-one

interaction. Afzal et al. (2019) note that the WT has

been trained based on best tutoring practices for each

student age group, though it remains challenging to

train the system to perform high-level soft human

skills required in tutoring.

While an AI-based tutor system cannot replace

human tutors' soft skills, it has the potential to

considerably impact children and young people

worldwide, particularly those with learning

difficulties or disabilities. The affordability and

availability of tutoring services for less privileged

students can be drastically improved with the

possibility of an AI-based tutoring system, leading to

better educational and social outcomes (idem, 2019).

2.3.4 Health Literacy and Health Outcomes

Some applications of AI systems are transversal to

health outcomes (including health literacy) and

educational outcomes (including media and

information literacies development). Recently, AI-

based algorithms have demonstrated the potential to

diagnose pediatric ophthalmological diseases early,

preventing blindness development. That is the case of

retinopathy of prematurity (ROP), which causes

retinal detachment, resulting in complete vision loss.

More than 30.000 children lose their vision from ROP

worldwide yearly (Li et al., 2021). AI algorithms have

been recently integrated into a telemedicine system

for ROP. They have demonstrated high accuracy in

the early detection more accurate than human

experimented examiners (2021). The ROP's early

diagnostic impact is vital because it allows early

treatment, preventing, retarding, or reducing total

vision loss. Developing these AI-based systems

prevents physiological damage and protects the well-

being of children at risk of ROP. Again, the cost of

access to this kind of technology will determine the

actual benefit it can have. We have not found research

that directly studies AI-based technologies on health

literacy development during childhood.

3 LEGAL & REGULATORY

FRAMEWORK

Due to AI technologies' vast social, economic, and

ethical implications, governments and many

organizations are setting guidelines for its

development and implementation. Through the

documental review, we found that legal and

regulatory frameworks are being built by multilateral

organizations, regional instances, the private sector,

and regional and government (Figure 1).

3.1 Trends

UNICEF Innovation created a Memorandum on AI

and the Rights of the Child, which provides policy

guidance for centered-child AI development. The

policy guidance includes nine requirements that

governments, policymakers, and businesses should

meet when developing technologies. However, the

regulatory framework and policies show inequalities

among regions. In Africa, there is a notable disparity

in access to the knowledge, data, education, training,

and human resources necessary to develop and utilize

AI. Latin American countries face challenges in

policymaking, capacity, and adequate resources but

recognize the potential to impact social well-being

positively. Developed countries focus their regulatory

framework and policies on the ethical use of AI,

digital safety, child protection, and data preservation.

(Barakina et al., 2021; Sibal & Neupane, 2020;

Cabrol et al., 2020; UNICEF Innovation, 2020; see

Table 4).

Table 4: Guiding UNICEF criteria for governments,

policymakers, and businesses in child-centered AI

development.

Support children's development

and well-being

Ensure inclusion of and for

children

Let AI help develop their full

potential

Include them and those around

them

Prioritize fairness and non-

discriminatory for children

Protect children's data and

privacy

AI must be for all children

Ensure their privacy in an

AI world

Ensure safety for children

Provide transparency,

explainability, and

accountability for children

Children need to be safe in

an AI world

Children need to know how

AI impacts them. Actors need to

be accountable for that

Empower governments and

businesses with knowledge of

AI and children's rights

Prepare children for present

and future developments in AI.

Actors must know what

children's rights are and uphold

them

If children are well prepared

now, they can contribute to

responsible Ai in the future

Create an enabling environment

Make it possible for all to contribute to child-centered AI

The Potential of Artiﬁcial Intelligence and Emerging Technologies for Digital Accessibility in Childhood Literacies: A Critical Review of

the Literature

107

4 IMPLICATIONS FOR SOCIETY

For better or worse, the disruptive effects of AI will

transform children's lives in ways we cannot yet

understand. Therefore, our collective actions on AI

today are critical for shaping an inclusive future that

children deserve.

Although AI holds great potential for advancing

education, health, and well-being, its widespread

implementation also presents significant challenges

and risks that must be addressed with equal

dedication and responsibility to ensure AI's safe and

ethical application in these fields. In the various

spheres of society today, Artificial Intelligence and

emerging technologies are already a reality,

especially in developed countries. The education

sector could not ignore the tools and technologies

available to improve educational processes.

The modern scientific literature presents

optimistic research results that validate these

technologies in specific population segments,

promoting accessibility in education (teaching and

learning) or health (diagnosis and follow-up).

However, unlike other technologies we have

mastered, the potential and applicability of AI and

emerging technologies raise some reservations

among the various stakeholders.

4.1 Childhood Education and Literacies

Research shows that AI technologies in education

have been implemented in three main directions:

training, research, and qualifying specialists to work

with AI (Tuomi, 2018). However, according to

UNICEF, there are risks of increasing inequalities

due to uneven access to technology, limited digital

skills, and inability to leverage its benefits, which can

exacerbate the gap in socio-economic and

technological spheres between low-income and high-

income countries (UNICEF, 2021b). Most studies on

the applicability of AI in childhood education occur

in countries with more robust economies, leaving

low-income countries at risk of being left behind.

To mitigate the risks of inequality, it is crucial to

ensure equitable access to technology and develop AI

systems that consider the characteristics and

requirements of the context in which they are being

applied (OECD, 2021). The development approach

should be cyclical to continuously evaluate

technologies' impacts and promote updates or

modifications to the strategies, system, and

algorithms (UNICEF, 2021b).

The digital divide is a significant problem, and AI

in education can either close or exacerbate the gap. In

developing countries, less internet connectivity and

digital literacy may limit the potential of AI solutions.

According to UNICEF, North America, and China are

predicted to gain the most from AI, while developing

countries in Africa, Latin America, and Asia may

experience more modest gains (UNICEF, 2021b).

To ensure the success of AI solutions in less

developed countries, tech companies and other

stakeholders must design solutions tailored to each

region's specific contexts and needs. Rather than

simply transferring solutions from technology-

intensive countries, solutions should be developed

based on a deep understanding of the region or culture

to increase the likelihood of adoption and success

(Vinuesa et al., 2020).

In developed countries, AI-enabled learning tools

can help children learn how to collaborate and

develop critical thinking and problem-solving skills

(UNICEF, 2021b). Customized learning experiences

provided by robust learning platforms can address

individual user needs. However, children growing up

in underprivileged environments may fail to seize the

opportunities they require to thrive in a world

increasingly reliant on artificial intelligence without

access to AI-enabled services. This may undermine

their capacity to exercise their citizenship in politics

or civic affairs, restrict their prospects of becoming

active "prosumers," and render them inadequately

prepared for future challenges in the job market.

Teachers can also benefit from AI-enabled tools

that help them quickly develop educational programs,

freeing time to focus on other classroom requirements

or individual student needs (UNICEF, 2021b).

However, the absence of such tools may compromise

the quality of teaching and content made available to

children.

Algorithm bias is a risk associated with AI in

education. Such bias may lead to predictions,

instructions, and analysis of patterns that neglect

certain data or present results that respond only to a

specific context. Algorithm bias can also compromise

the expected outcome for children with special

learning needs. In a non-regulated environment,

development bias may promote exclusion and

discrimination.

4.2 Childhood Health Outcomes

AI-based developments in childhood healthcare

services for children with dyslexia, autism, and

motor, visual, or speech limitations show remarkable

potential. AI technologies have a wide range of

health-related applications, from early diagnosis and

gap-closing in mental healthcare to increasing health

CSEDU 2023 - 15th International Conference on Computer Supported Education

108

and eHealth literacy. Multimodal learning analytics,

interactive learning environments, and new research

instruments enable researchers to track and evaluate

emerging conceptual capacity and generate a new

form of student-teacher-knowledge interaction

(Abrahamson et al., 2020).

However, concerns about disorders related to

blending education models, such as screen fatigue

and stress adaptation, and falling behind those ill-

equipped for digital learning still exist (OECD, 2021).

Any approach to education must consider the gains

and implications for learning and children's health, as

AI applications are dynamic systems that learn and

adapt to their environment and will continue to be part

of various aspects of our lives.

4.3 Implications for Parents and

Caregivers

AI is increasingly affecting the lives of children

directly and indirectly, making it necessary for

parents and caregivers to play a more active role in

monitoring and guiding their use of technology. The

UNICEF AI Parenting Guide encourages caregivers

to consider the potential impacts of AI systems on

their children's privacy and data and advocate for the

responsible use of AI in schools and other settings

(UNICEF, 2021a). Parents and caregivers can help

children better understand the pros and cons of AI, as

well as control their exposure to AI systems and

personal data collection.

AI technologies offer opportunities for children

with physical or cognitive conditions to receive

personalized education and support, but they also

present risks and challenges that must be addressed.

Multimodal learning analytics, for example, enable

researchers to track and evaluate students' emerging

conceptual capacity while formalizing their gestures

and actions in disciplinary formats and language

(Abrahamson et al., 2020). However, parents and

educators must be aware of potential issues such as

screen fatigue, stress adaptation, and falling behind

for those unprepared for digital learning (OECD,

2021). As AI systems evolve and become more

prevalent in all aspects of life, it is essential to balance

the potential gains with the implications for children's

health and learning.

4.4 Implications for Teachers and

Schools

The Covid-19 pandemic led to the sudden adoption of

digital learning, with schools closed and education

moving online (OECD, 2021). This acceleration of

the digital transformation of education has brought

about innovative approaches to learning, design, and

analysis instruments and technologies in education

(Abrahamson et al., 2020; Barakina et al., 2021;

OECD, 2021). However, it is essential to note that

technology is only a tool to improve the quality of the

educational process, not a replacement for the teacher

or school. The use of AI and emerging technologies

in the educational system should focus on improving

the interaction between various stakeholders and not

limit the ability of children to "think for themselves"

(Barakina et al., 2021).

The main issue with digitalization and AI in

education is how education responds adequately to

emerging societal and labor-market needs. Therefore,

it is necessary to ensure that the performance of

algorithms is not biased and that the data and

information entered do not exclude individualities

(OECD, 2021). In addition, the increasing importance

of skills that are more difficult to automate, such as

creativity, critical thinking, communication, and

collaboration, should be considered in using AI in

education (Vincent-Lancrin et al., 2019 in OECD,

2021). It is essential to consider that emerging

technologies are still new tools not yet sufficiently

mastered by teachers, students, and parents and that

their safety and reliability must be precisely

confirmed before their successful use (Barakina et al.,

2021).

4.5 Implications for Tech Companies

and AI Developers

AI systems rely on data as their primary commodity

and can make decisions and predictions without

human involvement (UNICEF, 2021b). While AI has

contributed to improving education, there are

concerns about how data is obtained, and patterns are

identified, especially when decisions are made about

people (UNICEF, 2021b). The design of AI solutions

should involve all stakeholders, consider

affordability, and adapt to the evolving needs of

children and the education system (OECD, 2021).

Furthermore, the purpose and motivation for the

development of AI in education must consider its

impact on children directly and indirectly and its

contribution to preparing children for the future. The

literature highlights the importance of considering the

context in which AI is applied, including national AI

strategies and regional AI infrastructures. To promote

updates or modifications to strategies, systems, and

algorithms, the development approach of tech

companies must be cyclical, continuously evaluating

technologies and their impacts (OECD, 2021).

The Potential of Artiﬁcial Intelligence and Emerging Technologies for Digital Accessibility in Childhood Literacies: A Critical Review of

the Literature

109

4.6 Implications for Multilateral and

Supranational Instances

AI systems are mostly embedded within digital

systems and hardware. For this reason, AI systems-

related discussions must also approach ethical, legal,

and digital ecosystem issues. While beneficial,

explicability and accountability are robust principles

specific to AI systems, protecting user privacy,

fairness, and inclusion are relevant for the whole

digital ecosystem. The absence of a guiding matrix or

multilateral guidance policies leads to unmeasured

actions by business groups in motivating, developing,

and implementing some technologies in education

and data appropriation.

The existing initiatives of transnational entities

are still very marked by a specific generalization, and

discourse is still very focused on the realities of

technologically advanced countries that are also the

ones that contribute the most and collaborate in

carrying out studies on AI and emerging

technologies.

AI applications in education and health care can

improve learning outcomes, health, and well-being.

However, we need regulations and safeguards that

ensure that AI systems are reliable, safe, and

trustworthy (UNICEF, 2021b). Society needs legal

frameworks that ensure that misappropriation of data

and violation of children's rights, guarantees, and

freedoms effectively fall within the legal framework

of each country, such as money laundering, tax

evasion, and other social issues.

4.7 Sustainable Development Goals

(SDGs) 2030

To date, there is a lack of published studies that assess

the extent to which AI might impact all aspects of the

17 Sustainable Development Goals (SDGs).

Although AI-enabled technology can act as a catalyst

to achieve the 2030 Agenda, it may also trigger

inequalities that may act as barriers ones (Vinuesa et

al., 2020).

Although the positive impacts outweigh the large-

scale negative, reported potential impacts of AI report

positive and negative repercussions on SDGs. A

recent report on the role of AI in achieving the SDGs

states that "AI can help achieve 134 targets across all

the goals'' (2020). We focus on the implications of AI

on SDG 4 quality education for this document.

Linking AI policies and strategies with the SDGs

can prioritize equity and inclusion and advance

children's development and well-being (Pedro et al.,

2016). However, regulatory oversight is essential to

enable the positive impacts of AI. Currently, there is

little or no oversight of global AI systems that

contribute more than those presented in this work.

The duality of AI impact is related to identifying

different needs and addressing appropriate responses.

However, it may also lead to additional qualification

requirements that do not exist, especially in

developing countries where cultural values often

provide the answers to such needs.

5 CONCLUSIONS

Artificial intelligence can revolutionize childhood

education and make it more accessible for children of

all backgrounds and abilities. However, it is crucial to

approach this technology cautiously and consider its

ethical implications.

AI-powered tutoring systems could be a powerful

tool for improving educational outcomes for children,

particularly in subjects like math and science.

However, it is imperative to ensure that these systems

are appropriately designed and monitored to avoid

reinforcing existing biases or perpetuating

inequalities.

Health literacy is a critical component of

childhood education, and AI has the potential to

improve health outcomes by providing personalized

recommendations and guidance. However, ensuring

that these systems are transparent and based on

accurate and reliable data is critical.

The legal and regulatory framework around AI in

childhood education is still evolving, and

policymakers must consider this technology's

potential risks and benefits. There is a need for clear

guidelines and standards to ensure that AI is used

ethically and responsibly.

While certain limitations and challenges are

associated with using AI in childhood education,

overall, the potential benefits of this technology are

significant. As we continue to explore the

possibilities of AI, it is vital to prioritize the needs and

well-being of children and ensure that this technology

is used to promote equity and inclusion.

6 LIMITATIONS

The objective of studying AI in childhood education

and legal and regulatory frameworks was too

ambitious, given time, resources, and thematic

literacy constraints. Combining the two topics limited

the exploration of AI's implications for society.

CSEDU 2023 - 15th International Conference on Computer Supported Education

110

Nevertheless, the review provided a preliminary

understanding of the implications of AI for childhood

education and accessibility from a scientific and

regulatory perspective. This review was conducted

before the release of GPTChat, a language model

developed by OpenAI, which could be a critical tool

for further research in this area.

ACKNOWLEDGMENTS

Portuguese National funds financially support this

work through FCT – Foundation for Science and

Technology, IP, under the project UIDB/05460/2020.

REFERENCES

Moore, R., Lopes, J. (1999). Paper templates. In

TEMPLATE'06, 1st International Conference on

Template Production. SCITEPRESS.

Smith, J. (1998). The book, The publishing company.

London, 2

edition.

Abrahamson, D., Nathan, M. J., Williams-Pierce, C.,

Walkington, C., Ottmar, E. R., Soto, H., & Alibali, M.

W. (2020). The Future of Embodied Design for

Mathematics Teaching and Learning. Frontiers in

Education, https://www.frontiersin.org/articles/10.

3389/feduc.2020.00147/full

Afzal, S., Dempsey, B., D'Helon, C., Mukhi, N., Pribic, M.,

Sickler, A., Strong, P., Vanchiswar, M., & Wilde, L.

(2019). The Personality of AI Systems in Education:

Experiences with the Watson Tutor, a one-on-one virtual

tutoring system. Childhood Education, 95(1), 44–52.

https://doi.org/10.1080/00094056.2019.1565809

Barakina, E. Y., Popova, A. V., Gorokhova, S. S., &

Voskovskaya, A. S. (2021). Digital Technologies and

Artificial Intelligence Technologies in Education.

European Journal of Contemporary Education, 10(2),

285–296.

Cabrol, M., González, N., Pombo, C., & Sánchez, R.

(2020). FAIr LAC: Adopción ética y responsable de la

inteligencia artificial en América Latina y el Caribe

(Nota Técnica IDB-TN-1839; Notas Técnicas Sector

Social). Interamerican Bank of Development.

https://publications.iadb.org/publications/spanish/docu

ment/fAIr_LAC_Adopci%C3%B3n_%C3%A9tica_y_

responsable_de_la_inteligencia_artificial_en_Am%C3

%A9rica_Latina_y_el_Caribe_es.pdf

Hartmann, N. S., & Aluisio, S. M. (2021). Automatic

lexical adaptation in brazilian portuguese informative

texts for elementary education. Linguamatica, 12(2), 3–

27.

Joy, J., Balakrishnan, K., & Sreeraj, M. (2019). SiLearn: An

intelligent sign vocabulary learning tool. Journal of

Enabling Technologies, 13(3), 173–187.

Li, J.-P. O., Liu, H., Ting, D. S. J., Jeon, S., Chan, R. V. P.,

Kim, J. E., Sim, D. A., Thomas, P. B. M., Lin, H., Chen,

Y., Sakomoto, T., Loewenstein, A., Lam, D. S. C.,

Pasquale, L. R., Wong, T. Y., Lam, L. A., & Ting, D.

S. W. (2021). Digital technology, telemedicine and

artificial intelligence in ophthalmology: A global

perspective. Progress in Retinal and Eye Research, 82.

https://pubmed.ncbi.nlm.nih.gov/32898686/

Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., &

Group, T. P. (2009). Preferred Reporting Items for

Systematic Reviews and Meta-Analyses: The PRISMA

Statement. PLOS Medicine, 6(7), e1000097.

https://doi.org/10.1371/journal.pmed.1000097

OECD. (2021). OECD Digital Education Outlook 2021:

Pushing frontiers with Artificial Intelligence,

Blockchain, and Robots. OECD publishing.

https://doi.org/10.1787/589b283f-en

Ouzzani, M., Hammady, H., Fedorowicz, Z., &

Elmagarmid, A. (2016). Rayyan—A web and mobile

app for systematic reviews. Systematic Reviews, 5(1),

210. https://doi.org/10.1186/s13643-016-0384-4

Park, C. S.-Y., Kim, H., & Lee, S. (2021). Do Less

Teaching, Do More Coaching: Toward Critical

Thinking for Ethical Applications of Artificial

Intelligence. Journal of Learning and Teaching in the

Digital Age, 6(2), 97–100.

Pedro, A. R. Amaral, O., & Escoval, A. (2016). Literacia

em saúde, dos dados à ação: Tradução, validação e

aplicação do European Health Literacy Survey em

Portugal. Revista Portuguesa de Saúde Pública, 34(3),

259–275. https://doi.org/10.1016/j.rpsp.2016.07.002

Saltman, K. J. (2020). Artificial Intelligence and the

Technological Turn of Public Education Privatization:

In Defence of Democratic Education. London Review

of Education, 18(2), 196–208.

See, A. R., & Advincula, W. D. (2021). Creating tactile

educational materials for the visually impaired and

blind students using ai cloud computing. Applied

Sciences (Switzerland), 11(16). https://www.mdpi.

com/2076-3417/11/16/7552

Shahamiri, S. R., & Thabtah, F. (2020). Autism AI: A New

Autism Screening System Based on Artificial

Intelligence. Cognitive Computation, 12(4), 766–777.

https://doi.org/10.1007/s12559-020-09743-3

Sibal, P., & Neupane, B. (2020). Artificial intelligence

needs assessment survey in Africa [Regional survey].

UNESCO. https://unesdoc.unesco.org/ark:/48223/pf

0000375410

Tuomi, I. (2018). The impact of Artificial Intelligence on

learning, teaching, and education. Policies for the future

(M. Cabrera, R. Vuorikari, & Y. Punie, Eds.).

Publications Office of the European Commission. Joint

Research Centre. https://data.europa.eu/doi/10.2760/

12297UNICEF

UNICEF (2021a). AI guide for parents (p. 5) UNICEF.

https://www.unicef.org/globalinsight/media/2336/file/

UNICEF-Global-Insight-AI%20guide%20for%20pare

nts-2021.pdf

UNICEF. (2021b). Policy guidance on AI for children 2.0

(Global Insight) [Working document]. UNICEF.

The Potential of Artiﬁcial Intelligence and Emerging Technologies for Digital Accessibility in Childhood Literacies: A Critical Review of

the Literature

111

https://www.unicef.org/globalinsight/media/2356/file/

UNICEF-Global-Insight-policy-guidance-AI-children-

2.0-2021.pdf.

Veinot, T. C., Mitchell, H., & Ancker, J. S. (2018). Good

intentions are not enough: How informatics

interventions can worsen inequality. Journal of the

American Medical Informatics Association, 25(8),

1080–1088.

Vincent-Lancrin, S., et al. (2019), Fostering Students'

Creativity and Critical Thinking: What it Means in

School, Educational Research and Innovation, OECD

Publishing, https://doi.org/10.1787/62212c37-en.

Vinuesa, R., Azizpour, H., Leite, I., Balaam, M., Dignum,

V., Domisch, S., Felländer, A., Langhans, S. D.,

Tegmark, M., & Fuso Nerini, F. (2020). The role of

artificial intelligence in achieving the Sustainable

Development Goals. Nature Communications, 11(33),

10. https://doi.org/| https://doi.org/10.1038/s41467-

019-14108-y

Wikipedia Collaborators. (2021). Artificial intelligence.

In Wikipedia. https://en.wikipedia.org/w/index.php?

title=Artificial_intelligence&oldid=1058701730.

CSEDU 2023 - 15th International Conference on Computer Supported Education

112