Why Professional Discretion Matters When Decision Support Crosses into Decision Shaping
This brief is offered in the spirit of inquiry rather than prescription. It extends prior edblogcast briefs by examining educator agency as a necessary condition for responsible judgment in AI-mediated educational systems.
Why this brief, and why now
In the previous brief, When Recommendations Shape Judgment, we examined how AI-mediated systems increasingly organize evidence, structure attention, and orient action through rankings, alerts, and recommended next steps. That analysis surfaced an important insight: even when humans retain formal decision-making authority, decision environments can subtly shape how judgment is exercised.
If judgment is exercised within environments influenced and structured by AI-enabled systems, then the question is no longer simply whether educators remain “in the loop.” The question is whether they retain meaningful control or professional agency — the capacity and opportunity to notice, interpret, contest, delay, adapt, or refuse action in light of contextual knowledge, professional values, and ethical responsibility (Alon-Barkat & Busuioc, 2023; Eteläpelto et al., 2013; Priestly et al., 2012, 2015; Seaver, 2019; Zeiser, 2024)
As AI becomes embedded in everyday educational workflows, educator agency cannot be assumed. It must be examined.
Who we mean by “educators”
In this brief, educators refers broadly to the professionals responsible for instructional, assessment, and learning-related decisions across educational systems. This includes, but is not limited to:
- Classroom teachers and instructional specialists
- School- and district-level leaders (e.g., principals, instructional coaches, curriculum directors)
- Assessment, data, and accountability professionals whose work shapes instructional signals and consequences
- Higher-education instructors and academic leaders operating within AI-mediated teaching and assessment environments
While these roles may differ in proximity to students and scope of responsibility, they share a common commitment: exercising professional judgment in contexts where AI increasingly influences what appears visible, relevant, and actionable.
Educator agency as a professional capacity
Educator agency is often framed in the literature as autonomy or freedom of choice (Eteläpelto et al., 2013; Priestley et al., 2015). Across education, governance, and the learning sciences, professional agency is conceptualized as a capacity rather than a personal trait — the ability to exercise discretionary judgment within institutional, technical, and normative constraints (Biesta et al., 2015; Pyhältö et al., 2014). This capacity is enacted relationally rather than individually and is influenced by policies, tools, accountability structures, and the time and resources available (Eteläpelto et al., 2013; Pyhältö et al., 2014).
In AI-mediated systems, this distinction matters. Educators may retain nominal authority while experiencing a reduction in practical agency if systems:
- predefine what counts as relevant evidence,
- structure workflows in ways that compress decision cycles and constrain opportunities for deliberation, or
- embed recommendations as defaults, norms, or expected next steps within operational processes (Williamson & Piattoeva, 2019; Seaver, 2019).
Under these conditions, professional judgment may be retained in principle but constrained in practice.
How AI reshapes the conditions for educator agency
AI systems do not simply provide information; they shape the decision environments in which professional judgment is exercised. Research on algorithmic governance and decision support identifies several design- and system-level mechanisms through which educator agency may be constrained, even when AI tools are framed as “supportive” (Alon-Barkat & Busuioc, 2023; Seaver, 2019; Williamson & Piattoeva, 2019).
1. Agency can be narrowed through attention structuring
Dashboards, alerts, and rankings foreground selected indicators while rendering others less visible. Over time, this can progressively narrow the scope of professional noticing — what educators attend to, question, or treat as consequential (Seaver, 2019; Williamson & Eynon, 2020). When certain metrics become established focal points, alternative interpretations and locally relevant considerations may be sidelined, not through mandate, but through design.
2. Agency can be compressed through pace
AI-enabled systems often promise efficiency and responsiveness. While speed can be beneficial, system-designed decision cycles that prioritize rapid response may reduce opportunities for reflection, consultation, and deliberation — particularly in high-stakes or ambiguous situations (Dieterle et al., 2024; Holmes et al., 2019). When rapid response becomes normalized, the professional discretion to pause or defer action can diminish.
3. Agency can be reframed through perceived legitimacy
Algorithmic outputs often carry a perceived objectivity or technical authority, especially when presented with numerical precision or confident language. Studies of automation bias and algorithmic decision-making show that professionals may over-rely on such outputs or feel pressure to justify deviations from them, even when they recognize limitations or contextual misalignment (Alon-Barkat & Busuioc, 2023; Wang, 2024).
In these ways, AI does not replace educators’ judgment, but it can systematically reshape the conditions under which that judgment is exercised.
Why “human-in-the-loop” does not ensure professional agency
Policy and governance discussions frequently emphasize keeping humans “in the loop” as a safeguard against automation-related risks (NIST, 2023; UNESCO, 2023). While necessary, this framing alone is insufficient.
Being in the loop does not ensure professional agency if educators are expected to act on outputs they did not help define, cannot meaningfully interrogate, or are institutionally constrained from contesting. When responsibility is assigned without corresponding discretion, agency is diminished rather than protected (NIST, 2023; UNESCO, 2023).
Emerging international guidance increasingly recognizes this distinction. UNESCO’s work on AI and education emphasizes that safeguarding educator agency requires attention not only to oversight, but to how AI systems are designed, introduced, and governed within everyday professional practice (UNESCO, 2023). Similarly, OECD underscores that professional agency is essential for navigating uncertainty and complexity in environments involving technology, particularly where judgment, interpretation, and contextual knowledge remain central to decision-making (OECD, 2025).
Educator agency as a condition for responsible innovation
Across edblogcast’s Foundations Series, responsible innovation was framed not as a technical achievement, but as alignment among human judgment, ethical accountability, and system design. Educator agency is a necessary condition for sustaining that alignment in practice.
When educators retain meaningful agency, AI-enabled tools and systems can function as intended — supporting rather than substituting for professional judgment. In such contexts, educators are able to:
- engage AI outputs as inputs for considerations, rather than directives for action,
- interpret and adapt recommendations in light of local contexts, learner needs, and professional knowledge, and
- remain accountable for decisions while exercising discretion about how and when system outputs inform practice.
When agency is constrained, the risk is not limited to uneven or ineffective implementation. It is a subtle redistribution of responsibility away from educators’ professional judgment and toward system-level processes that operate with assumptions, priorities, and decision logic that are not fully visible or contestable, even as educators continue to be held publicly accountable for outcomes (Williamson & Piattoeva, 2019).
Closing Reflection
This brief does not argue against AI in education, nor does it suggest that educator agency should remain unchanged. Instead, it invites a focused question for the phase of educational innovation:
What does it mean to exercise professional agency when decision environments are increasingly shaped by AI, and what system conditions are required to ensure that such agency is substantive rather than merely nominal?
Engaging this question is a prerequisite not only for responsible AI use, but for sustaining the developmental, relational, and ethical dimensions of education that depend on professional judgment.
Reflective questions
What conditions in your AI-supported workflows most enable the thoughtful exercise of professional judgment (e.g., time to deliberate, opportunities to pause, defaults that invite review, or space for consultation)?
Where do AI tools in your setting create space for reflection, contextual interpretation, or collaborative sense-making before decisions are acted upon?
Which aspects of decision-making remain clearly supported by educator discretion when AI systems are involved, and how are those enabling conditions intentionally designed or sustained?
Join this post’s Discussion & Reflection below.
Continue the Inquiry and Reflection
Members receive resources related to each brief and podcast.
About this brief
edblogcast explores how evidence, ethics, and innovation can evolve together in education’s age of AI. Each brief invites reflection and dialogue across educators, researchers, and innovators — not to prescribe solutions, but to deepen shared understanding of how responsible innovation takes shape in practice.
Together, these explorations aim to bridge perspectives — from systems design to classroom realities — and to support a professional community that learns through inquiry, collaboration, and care.
Join the conversation at edblogcast.com
→ Subscribe, share your perspective, or submit a study.
This space is currently unavailable as updates are underway.
It will be restored soon.
For now, you’re welcome to contact us at team@edblogcast.com
© edblogcast.com. All rights reserved.
Content provided for informational purposes and professional reflection. Materials are intended to support inquiry, dialogue, and thoughtful engagement, not to prescribe specific practices or policies.
References
Alon‑Barkat, S., & Busuioc, M. (2023). Human–AI interactions in public sector decision‑making: Automation bias and selective adherence to algorithmic advice. Journal of Public Administration Research and Theory, 33(1), 153–170.
Biesta, G., Priestley, M., & Robinson, S. (2015). The role of beliefs in teacher agency. Teachers and Teaching, 21(6), 624–640.
Dieterle, E., Dede, C., & Saxberg, B. (2024). AI-enabled decision-making in education: Ethics, evidence, and professional responsibility. Harvard Graduate School of Education.
Emirbayer, M., & Mische, A. (1998). What is agency? American Journal of Sociology, 103(4), 962–1023.
Eteläpelto, A., Vähäsantanen, K., Hökkä, P., & Paloniemi, S. (2013). What is agency? Conceptualizing professional agency at work. Educational Research Review, 10, 45–65.
Holmes, W., Bialik, M., & Fadel, C. (2019). Artificial intelligence in education: Promise and implications for teaching and learning. Center for Curriculum Redesign.
National Institute of Standards and Technology. (2023). Artificial Intelligence Risk Management Framework (AI RMF 1.0) (NIST AI 100‑1).
OECD. (2025). Section 4: Teacher agency to lead curriculum change. Teaching Compass Concept Note. OECD.
Priestley, M., Biesta, G., & Robinson, S. (2015). Teacher agency: What is it and why does it matter? In J. Evers & R. Kneyber (Eds.) Flip the System: Changing Education from the Ground Up. London. Routledge.
Priestley, M., Edwards, R., Priestley, A., & Miller, K. (2012). Teacher agency in curriculum making: Agents of change and spaces for manoeuvre. Curriculum Inquiry, 42(2), 191-214.
Pyhältö, K., Pietarinen, J., & Soini, T. (2014). Comprehensive school teachers’ professional agency in large-scale educational change. Journal of Educational Change, 15, 303-325.
Seaver, N. (2019). Knowing algorithms. Social Studies of Science, 49(3), 412–422. https://doi.org/10.1177/0306312719827617
UNESCO. (2023). Guidance for generative AI in education and research. UNESCO Publishing.
Wang, J. (2024). Algorithmic governance in education: Decision-making, discretion, and responsibility. Educational Administration Quarterly, 60(1), 3–30. https://doi.org/10.1177/0013161X231201234
Williamson, B., & Eynon, R. (2020). Historical threads, missing links, and future directions in AI in education. Learning, Media and Technology, 45(3), 223–235. https://doi.org/10.1080/17439884.2020.1798995
Williamson, B., & Piattoeva, N. (2019). Objectivity as standardization in data-scientific education policy. Learning, Media and Technology, 44(1), 64–76. https://doi.org/10.1080/17439884.2018.1556215
Zeiser, J. (2024). Owning decisions: AI decision-support and the attributability-gap. Science and Engineering Ethics, 30(7)
Discussion & Reflection
This space is for shared reflection and inquiry. Rather than quick reactions, we invite thoughtful contributions that extend the questions raised in the brief.
This space is currently unavailable as updates are underway.
It will be restored soon.
For now, you’re welcome to contact us at team@edblogcast.com
