Full List of AI Ideas

We want to hear from you on what you think are the most promising ideas for AI in Education. You can filter in three ways – where the idea intersects with the education system, what type of AI tools would be used and the overall purpose of the idea.

Please upvote/ downvote ideas by clicking the up/ down arrows next to the idea.

Ideas for AI in Education

We want to hear from you on what you think are the most promising ideas of AI that can help improve teaching and learning - you can vote by clicking the up arrow next to the idea. You can filter by category, by 'What does it do?', by AI tool or search for keywords.
Category What is the issue? Solution What does it do? What kinds of AI tools can be used? Votes
Continuous Professional Development, Pre-service Teacher Education Teachers do not receive enough support from colleagues and mentors. AI can help with analysis of teachers’ skills, weaknesses and interests and can support formation of effective collaborative working groups or support networks. Analytics Large Language Models, ML: Classification, ML: Clustering, ML: Dimensionality Reduction, ML: Regression, NLP: Generation
5
Continuous Professional Development, Positive Learning Environment Limited opportunities for feedback in education systems. Support feedback collection and analysis to inform improvements e.g. teachers feedback about professional development or students about teachers. Analytics Large Language Models, NLP: Generation, NLP: Text Classification
7
Assess & Review, Leadership of T&L, Teacher Learning Teachers do not often have access to mentors to do observations and give them feedback on classroom practice. Analyse lesson recordings using AI to provide feedback to teachers on their classroom practice and promote reflective practice through questioning. Digitisation, Management and use of education information Audio: ASR, Audio: Classification, Audio: Text2Speech, Large Language Models, ML: Classification, ML: Clustering, ML: Dimensionality Reduction, ML: Regression, NLP: Generation
10
Continuous Professional Development, Teacher Learning Teachers in remote areas may lack access to peersupport or mentors for pedagogical and content knowledge and skills. Chatbots respond to teacher/ teacher trainer questions and guide reflection using voice or text. 1-2-1 support, Personalisation Audio: ASR, Audio: Text2Speech, Large Language Models, NLP: Generation, NLP: Question
12
Continuous Professional Development, Pre-service Teacher Education The quality of teacher development programmes is variable and there is often a disconnect with pre-service teacher education. Use AI to develop resources and syllabuses for both pre- and in-service teacher education. Content creation & adaptation Large Language Models, NLP: Generation
10
Pre-service Teacher Education Pre-service teacher education curricula do not always align either to current professional practice or to current government policy. AI can help review existing pre-service teacher curricula to ensure alignment with best practice and national policy. Content creation & adaptation Large Language Models, NLP: Generation, NLP: Text Classification
12
Continuous Professional Development, Data Quality & EMIS, Leadership of T&L There is limited data analysis on student assessments at a national level so targeting support to underperforming schools is challenging. Analyse school performance data and use AI to group schools to help inform targeting of training and support. Analytics, Management and use of education information ML: Classification, ML: Clustering, ML: Dimensionality Reduction
4
Leadership of T&L Administrators need more information on classroom practice to facilitate improvement. Using planning documents and classroom recordings, provide summaries of classroom proceedings and content to give school leaders information on teacher practice and curriculum coverage. Management and use of education information Audio: ASR, Large Language Models, NLP: Generation, NLP: Summarisation, NLP: Text Classification
3
Data Quality & EMIS, Enabling Policies Governments need more data analysis to understand whether educational interventions are successful. Aggregation and analysis of data relating to education intervention programmes to understand their effectiveness. Analytics, Management and use of education information Large Language Models, ML: Classification, ML: Clustering, ML: Dimensionality Reduction, ML: Regression, NLP: Generation, NLP: Summarisation
5
Leadership of T&L Administrative tasks can be repetitive and time consuming. Generating suggestions for administrative tasks such as drafts of letters and contracts, talking points for meetings or questions for interviews can save time. Content creation & adaptation Large Language Models, NLP: Generation, NLP: Text Classification
2
Leadership of T&L School leaders have limited data to help them understand teacher performance. Analyse any combination of student performance data, data from classroom observations and teacher assessments to provide information on teacher performance. Analytics, Management and use of education information Audio: ASR, Large Language Models, ML: Classification, ML: Clustering, ML: Dimensionality Reduction, ML: Regression, NLP: Generation
0
Pre-service Teacher Education, Recruitment Better assessment of teaching candidates is needed to ensure standards are maintained and development tailored usefully. Designing and analysing assessments of pre-service teachers' personal and professional qualities to ensure support can be tailored effectively. Analytics, Management and use of education information Large Language Models, NLP: Generation, NLP: Text Classification
1
Data Quality & EMIS, Recruitment Teachers are placed in schools where they are unsuitable due to personal, cultural, or linguistic factors. Analysis of staff retention alongside assessments of teachers' personal and professional qualities provides information to help inform hiring or posting decisions. Analytics, Management and use of education information Large Language Models, ML: Classification, ML: Clustering, ML: Dimensionality Reduction, NLP: Generation, NLP: Text Classification
3
Data Quality & EMIS, Leadership of T&L, School Inputs Budgeting and planning require the synthesis of lots of data sources to evaluate past and predict future performance. Synthesise multiple data sources in reports to support budgeting and planning processes. Analytics, Management and use of education information ML: Classification, ML: Clustering, ML: Dimensionality Reduction, ML: Regression
2
Leadership of T&L Timetabling is a time-intensive task for school leaders who could better use this time. Draft school timetables for review by school leaders. Content creation & adaptation Large Language Models, NLP: Generation
-4
Data Quality & EMIS, Leadership of T&L, Student and family factors School dropout for academic or social reasons is a big problem for many schools. Analyse student performance and wellbeing data to identify students at risk of dropout. Analytics, Inclusion, Management and use of education information ML: Classification, ML: Clustering, ML: Dimensionality Reduction
9
Leadership of T&L, Student and family factors Many schools have little or no regular contact with parents. Generate electronic messages to parents automatically including students' report cards with explanations or suggestions for continued learning. Analytics, Content creation & adaptation, Management and use of education information Large Language Models, NLP: Generation
-5
Assessment Tools & Processes, Data Quality & EMIS, Enabling Policies School leaders and policy makers struggle to use data due to low levels of data literacy. Current data dashboards rely on charts and not interpretations. AI can help interpret charts in dashboards at school level to allow for local planning. Analytics, Management and use of education information ML: Classification, ML: Clustering, ML: Dimensionality Reduction, ML: Regression, MM: Image2Text, NLP: Generation, NLP: Summarisation
3
Leadership of T&L, Quality Curriculum Writing curricula is expensive and time-consuming. Review and draft changes to curricula at school or national level. Content creation & adaptation Large Language Models, NLP: Generation, NLP: Text Classification
-1
Plan Alignment between planning materials and curricula is needed to improve coverage in schools. Suggest revisions and additions after comparing planning documents to approved curriculum and policy documents. Content creation & adaptation Large Language Models, NLP: Generation, NLP: Summarisation, NLP: Text Classification
8
Assessment Tools & Processes, Student Learning Students do not have enough support to prepare well for examinations. Generate revision questions and provide adaptive platforms to support examination preparation. Personalisation Large Language Models, ML: Classification, ML: Clustering, ML: Dimensionality Reduction, ML: Regression, NLP: Generation, NLP: Text Classification
7
Assess & Review Ability to assess oral language ability effectively is either limited or time-consuming. Tools for detecting voice and assessing proficiency in a spoken language. Digitisation Audio: ASR, Large Language Models, ML: Classification, ML: Clustering, ML: Dimensionality Reduction, NLP: Generation
0
Assess & Review Assessment results are not well used to inform teaching or help struggling students catch up. Assessment analysis used to identify learning gaps, identify areas for support and make personalised recommendations for future learning and effective remediation. Analytics, Personalisation Large Language Models, ML: Classification, ML: Clustering, ML: Dimensionality Reduction, ML: Regression, MM: Image2Text, NLP: Generation
0
Assessment Tools & Processes Assessments are often poorly assigned because of a lack of knowledge about which questions or tasks will be most effective. Evaluate the validity and effectiveness of assessment tasks. Analytics, Content creation & adaptation Large Language Models, ML: Classification, ML: Clustering, ML: Dimensionality Reduction, ML: Regression, NLP: Generation
1
Assessment Tools & Processes Assessments are not always effective, due to bias or a lack of connection with learning. Check existing assessments for alignment with curriculum and for bias. Content creation & adaptation Large Language Models, NLP: Generation, NLP: Text Classification
2
Assess & Review Current ways of presenting key assessment data lack predictive capability so teachers are not able to plan effectively. AI can analyse student assessments and develop reports on achievement for teachers that provide predictions to support planning. Analytics, Management and use of education information ML: Classification, ML: Clustering, ML: Dimensionality Reduction
13
Assess & Review, Student Learning Large class sizes make marking written classroom assessments time-consuming and limit student feedback. AI can facilitate automatic marking of handwritten assessments and generation of personalised student feedback. Digitisation, Personalisation Large Language Models, Large Multi-Modal Models, ML: Classification, MM: Image2Text, NLP: Generation
4
Assessment Tools & Processes Classroom assessments often fail to capture progress of those who cannot access written tests. Voice-AI can transcribe a verbally administered assessments. Digitisation, Inclusion Audio: ASR
14
Assessment Tools & Processes Marking of large-scale assessments is labour-intensive and prone to human error. AI can facilitate automatic marking of short answer written assessments. Analytics, Digitisation Large Language Models, ML: Classification, MM: Image2Text, NLP: Generation, NLP: Text Classification
14
Assessment Tools & Processes Screening for reading difficulties is complex and time consuming. AI diagnostic tools can listen and analyse student responses in verbal reading assessments and generate recommendations for inclusion. Analytics, Inclusion Audio: ASR, Large Language Models, ML: Classification, ML: Clustering, ML: Dimensionality Reduction, NLP: Generation
0
Assessment Tools & Processes Written assessments are challenging for students with reading difficulties and limits data availability on their progress. AI can read exam questions out to students. Inclusion Audio: Text2Speech
5
Assessment Tools & Processes Generation of meaningful assessments related to learning objectives is a skill requiring professional training and practice. AI can create assessment tasks that link to lesson plans and the curriculum. Content creation & adaptation Large Language Models, NLP: Generation, NLP: Text Classification
16
Plan Teachers do not receive enough support from colleagues and mentors to design cross-curricular activities. Chatbot leads conversations with teachers from different disciplinary areas to structure lesson plans to support delivery of cross-curricular learning activities using available resources. Content creation & adaptation Audio: ASR, Audio: Classification, Large Language Models, NLP: Generation, NLP: Text Classification
0
Plan, Quality Teaching/ Learning Materials Many schools have limited resources, but teachers lack creative ways to supplement these with additional ones in their local community. AI tools can analyse resources and suggest how these can be relevant to curriculum. Content creation & adaptation Large Language Models, Large Multi-Modal Models, ML: Classification, ML: Clustering, ML: Dimensionality Reduction, ML: Regression, MM: Image2Text, NLP: Generation, NLP: Text Classification
2
Plan Teachers use a limited range of pedagogical strategies, which do not often include effective but complex approaches such as project-based learning. Use of AI to analyse opportunities for project-based learning and suggest possible cross curricular links. 1-2-1 support, Personalisation Large Language Models, NLP: Generation, NLP: Text Classification
0
Quality Teaching/ Learning Materials Limited access to good quality learning materials for students in areas of limited connectivity. Access to AI personalised learning within offline devices. Content creation & adaptation, Personalisation Large Language Models, ML: Classification, ML: Clustering, ML: Dimensionality Reduction, ML: Regression, NLP: Generation
1
Assess & Review, Plan Teachers are not able to monitor or support students working in groups because of large class sizes. Record group work and use AI to analyse collaborative discussion to generate suggestions for future teaching and support. Analytics, Digitisation, Management and use of education information Audio: ASR, Audio: Classification, Audio: Text2Speech, Large Language Models, ML: Classification, ML: Clustering, ML: Dimensionality Reduction, ML: Regression
11
Quality Teaching/ Learning Materials Learning resources are often only text based and lack multi-media features. AI can convert text-based learning resources to audio or video formats to improve accessibility. Content creation & adaptation, Enhancing educational experience, Inclusion Audio: Text2Speech, MM: Text2Image, MM: Text2Video
17
Quality Teaching/ Learning Materials Images in learning materials are often not culturally or geographically relevant so are hard for students to relate to. AI can generate culturally relevant images relating to specific text in teaching resources. Content creation & adaptation CV: Image2Image, Large Multi-Modal Models, MM: Text2Image
18
Plan, Quality Teaching/ Learning Materials Teachers need the support of well trained and experienced colleagues to review existing learning resources. AI can review existing learning resources and suggest additions and improvements. Content creation & adaptation Large Language Models, NLP: Generation, NLP: Text Classification
17
Quality Teaching/ Learning Materials Not all schools have good quality resources that link learning objectives to the local cultural context. AI can create learning resources that are relevant to the local cultural context. Content creation & adaptation NLP: Translation
43
Language of Instruction, Quality Teaching/ Learning Materials Teaching younger students in their mother tongue is desirable. However, resources in local languages are limited. Translate resources into local languages and create audio content. Content creation & adaptation, Inclusion NLP: Translation
23
Plan Varied levels and quality of teacher training programmes means teachers lack the relevant pedagogical and content knowledge to access teaching materials. AI can provide concise summaries of relevant teaching resources or information about ways to teach in accessible languages. Content creation & adaptation Large Language Models, NLP: Generation, NLP: Summarisation
8
Plan Some teachers are not familiar with current best practices resulting in poor quality planning and learning activities. AI can create lesson plans informed by best practice. Content creation & adaptation Large Language Models, NLP: Generation
20
Teacher Learning Teachers often lack conversational discussions to explore ideas and develop their skills. AI Chatbots used by teachers as a talking partner to develop specific knowledge or skills. 1-2-1 support, Personalisation Audio: ASR, Audio: Text2Speech, Large Language Models, NLP: Conversational, NLP: Generation
11
Language of Instruction, Plan Teachers may not be native speakers of the official language of instruction, meaning accessing planning materials is hard. AI translation of teaching materials to improve access. Content creation & adaptation NLP: Translation
36
Student Learning Due to culture, displacement or Special Educational Needs or Disabilities (SEND), some students have difficulty understanding or interpreting images. Describe and explain images in accessible and contextually relevant terms. Content creation & adaptation, Enhancing educational experience, Inclusion CV: Image Classification, Large Multi-Modal Models, MM: Image2Text, NLP: Generation
6
Student Learning, Teach Opportunities for collaborative work are limited because of large class sizes. AI can group children into their ability groups for Teaching at the Right Level (TaRL) and then AI models can analyse collaboration and guiding discussion through questioning and feedback to learners. Enhancing educational experience Audio: ASR, Audio: Classification, Audio: Text2Speech, Large Language Models, ML: Classification, ML: Clustering, ML: Dimensionality Reduction, ML: Regression, NLP: Generation
1
Quality Teaching/ Learning Materials Absenteeism means students miss out on important learning. AI can provide summaries of classroom proceedings and content to enable absent students to catch up. Content creation & adaptation, Inclusion Audio: ASR, NLP: Generation, NLP: Summarisation, NLP: Text Classification
1
Quality Teaching/ Learning Materials, Student Learning Learning resources are limited mainly to textbooks and other printed material, which is passive rather than interactive. Augmented reality tools to enhance the properties of existing learning resources. Enhancing educational experience CV: Depth, CV: Object, CV: Segmentation
-5
Quality Teaching/ Learning Materials, Student Learning Long or dated texts can be inaccessible for students. AI can summarise learning content in an accessible format for students. Content creation & adaptation, Enhancing educational experience Large Language Models, NLP: Generation, NLP: Summarisation
17
Student Learning Students often lack conversational discussions due to large classes and rote learning. AI Chatbots used by students as a talking partner to develop specific knowledge or skills. 1-2-1 support, Personalisation Audio: ASR, Audio: Text2Speech, Large Language Models, NLP: Conversational, NLP: Generation
14
Student Learning Students often lack the necessary support for learning, due to large classes, absent teachers, or working parents. AI Chatbots respond to student questions about challenges with subject matter using voice or text. 1-2-1 support, Personalisation Audio: ASR, Audio: Text2Speech, Large Language Models, NLP: Generation, NLP: Question
21
Language of Instruction, Student Learning Students with difficulty writing are unable to show their understanding in other curriculum areas. AI can transcribe/translate pupil speech to write their assignments. Inclusion Audio: ASR, NLP: Translation
7
Student Learning In large classes, it is not possible for teachers to provide personalised feedback to all students. AI can provide individualised feedback on students' work. Personalisation Large Language Models, NLP: Generation, NLP: Summarisation, NLP: Text Classification
9
Quality Teaching/ Learning Materials, Student Learning In large classes, it is not possible for teachers to provide adequately differentiated tasks to all students. AI can assess students' learning levels and then can generate tasks of appropriate challenge. Analytics, Personalisation Large Language Models, ML: Classification, ML: Clustering, ML: Dimensionality Reduction, NLP: Generation
74
Back to Education System Diagram

How We Have Categorised

AI Tools

AI Tool Categories • Definitions

While developing our AI ideas list we’ve also put together a list of the AI tools that might be used to solve the issues set out in each idea. The aim is to help us identify and prioritise what tech to develop and to understand the problems we want to solve.

See AI Tools List

What Does It Do?

Function • Purpose

Some of the AI applications are performing similar functions (even though serving different parts of the Teaching and Learning Framework). For example, AI chatbots could be used by teachers to support Teacher Learning, or by students to support Student Learning. We’ve put together eight functions that AI tools perform to solve the education issues identified. The aim is to help us group similar ideas together for ease of navigation.

See Functions List

Have We Missed Anything?

You can suggest AI ideas for inclusion on the platform – if you think we’ve missed something please add your suggestions using the form via the link below.

Sign Up

Join our mailing list to keep up to date with news and events.

Community          Knowledge          FAQ
          Privacy Policy

AI-for-Education.org was set up by Fab Inc. in partnership with Team4Tech. We are grateful to the Bill & Melinda Gates Foundation and the Jacobs Foundation for their support.

Powered by FabData.IO