Exploring Teachers' Perceptions of Artificial Intelligence as a Tool to Support their Practice in Estonian K-12 Education

University learning analytics / education technology research (Estonia)

Estonia

Researchers administering survey and analyzing responses

• In-school (K–12)

Online or distributed survey of in-service K-12 teachers

Single-wave perception survey (analysis reported in journal article)

n = 140 Estonian K-12 teachers

Not intervention-specific; teachers reflect on AI-enhanced tools (ITS, analytics, authoring aids, etc.)

• Self-reported perceptions may exceed actual classroom AI fluency
• Single national context with strong digital infrastructure (limits generalization)
• Rapid AI tool change since 2021 may shift teacher familiarity
• Survey does not measure student learning outcomes directly

K-12 students indirectly (teachers as respondents about practice)

Uneven prior use of AI classroom tools across schools

Not required for teacher respondents; AI framed as practice support

• Characterize how Estonian teachers perceive AI as support for teaching and their expectations
• Surface perceived workplace challenges (“superpowers” metaphor) relevant to AI design
• Relate findings to FATE and participatory design of trustworthy classroom AI

• Highlight multilingual content access/adaptation as socio-cultural design requirement
• Inform professional development for ethical, transparent AI adoption

• Not a randomized trial of a specific AI product
• Not classroom observation of AI use frequency coded longitudinally
• Not a student achievement study

General AI-in-education affordances (ITS, analytics, personalization, assessment support—teacher-referenced)

• Tutor
• Automation tool
• Evaluator

Estonian education context with emphasis on multilingual teaching support

• Teachers envision AI amplifying planning, awareness of student progress, and adaptive scaffolding
• Metaphor elicitation aims to capture latent needs without locking to existing vendors

• FATE framing: fairness, accountability, transparency, and ethics must be explicit in procurement and design
• Mitigate algorithmic bias especially when models inform high-stakes judgments
• Interpretability and explainability for teacher trust and appropriate override
• Stakeholder communication with parents, unions, and policymakers in accessible terms

• Ground study in Estonian digital-education leadership and prior low AI-awareness findings
• Deploy structured survey probing AI perceptions and “superpower” teaching challenges
• Analyze responses qualitatively/quantitatively and connect to FATE design principles
• Derive implications for AI-enhanced tools and teacher PD

• Teachers remain accountable for instructional decisions; AI augments awareness and workload support
• Systems should expose limits and invite teacher judgment rather than opaque automation

• Use participatory methods when moving from survey insights to co-designed tools
• Pair technical rollout with ethics literacy for faculty

• Limited knowledge of how AI could concretely support daily practice
• Uncertainty about school policies promoting AI in education
• Need for multilingual content access and adaptation in local context
• Bridging gap between digital readiness indices and classroom-level AI integration

• Borrowed “teacher superpowers” metaphor from prior HCI/AIED work to elicit needs without vendor lock-in
• Cross-walked results with Aiken & Epstein-style AI design principles under FATE

• Teachers largely perceive AI as an opportunity despite knowledge limits
• Rich qualitative signals on desired supports and socio-cultural constraints

• Evidence supports need for structured PD and trustworthy tool design before scale adoption
• Findings motivate participatory co-design with Estonian teachers as next step

Authors connect results to ethical AI deployment and future participatory design of learning environments aligned with teacher-stated needs.

• Survey data require anonymization and secure storage consistent with institutional ethics
• Future AI systems that analyze teacher or student behavior need strict purpose limitation
• Transparency obligations when models inform administrators about classrooms
• Fairness reviews when AI supports multilingual learners to avoid deficit bias

• Post assessment
• Activity documentation
• Practitioner observation

• Longitudinal studies linking teacher AI literacy PD to classroom adoption
• Design interventions testing FATE-checklisted AI dashboards in Estonian or similar systems

• Teacher cognition and AI acceptance
• Trustworthy / ethical AI in education
• Participatory design in K-12

• Completed

K-12 (teacher-focused study)

Estonian formal schools

Teacher support / awareness / personalization (anticipated)

Survey-informed design implications

Medium

Medium (future analytics on teaching)