Teachers have always relied on professional judgment to know when to step in, when to step back and when to let students work through challenges on their own. Now, AI is adding another layer to those decisions. Educators must navigate not only how technology can support their own work, from lesson planning to assessment, but also how and when students should use it to support their learning.

As AI becomes more embedded in education, the conversation needs to move beyond which tools teachers and students are using to whether those tools are actually improving learning. The challenge is not simply knowing how to use AI, but recognizing when it adds value, when it risks getting in the way and when a different approach better serves students. In an AI-enabled classroom, knowing when not to use technology is becoming just as important as knowing how to use it.

Can teachers use AI tools effectively? Do they understand how AI works? Discussions around AI readiness are important, but they represent only one part of the story.

ETS’s AI Competencies for Teaching research presents AI readiness as multidimensional, bringing together pedagogical knowledge, content understanding, technological skills, ethical awareness, professional judgment and a commitment to continuous learning. AI competency is not treated as a standalone technical skill; it is viewed as a combination of professional knowledge and human judgment applied in complex instructional settings.  skill; it is viewed as a combination of professional knowledge and human judgment applied in complex instructional settings.

At its core, AI competency requires teachers to answer three simple but important questions:

• Should AI be used?
• How should I use it?
• When should AI not be used?

These questions shift the conversation away from technology itself and toward the professional judgments required to use it effectively in teaching. Competency becomes visible not only through the ability to operate AI tools, but through the quality of the decisions educators make about whether, when and how those tools can support learning. AI becomes more embedded in education, the human lens remains a critical layer in ensuring technology is serving learning, rather than shaping it.

A new dimension of professional judgment

Teaching has always depended on professional judgment and that will not change as this new technology becomes more integrated into the classroom. As educators navigate the opportunities and challenges of AI, their ability to determine when, how and whether to use it becomes another important dimension of effective teaching. ETS research suggests that choosing not to use AI in a particular context can itself demonstrate professional judgment and AI competency.

Consider critical thinking, one of the most important skills educators seek to develop. AI may accelerate a task, but it can also prevent students from engaging in the productive struggle that often leads to deeper understanding. Similarly, in a writing assignment designed to develop students’ voice and communication skills, AI may be valuable for revision or feedback, but excessive reliance on generated content could undermine the very skills the assignment is intended to strengthen. The same principle applies to the human interactions at the heart of teaching.

Building trust with students means identifying gaps in understanding or responding to social and emotional needs. This is where human connection outweighs technological efficiency.

As AI systems advance, teachers face more decisions about where those capabilities belong in teaching and learning. There are risks associated with improper use of AI, making human expertise increasingly more important. Professional judgment will remain central to effective teaching, regardless of how sophisticated the technology becomes.

Teachers will increasingly serve as evaluators of when AI enhances learning, guardians of student agency and decision-makers responsible for maintaining the human relationships that support student development.

If educational systems want to prepare teachers for an AI-rich future, they may need to rethink how readiness is defined and measured.

Too often, readiness is framed in terms of access, adoption or technical proficiency. But those indicators tell us little about how effectively AI is being integrated into learning.

ETS research suggests a different approach.  AI competencies should be understood as developmental, growing through experience and practice. They grow through authentic classroom practice, reflection, feedback and professional learning opportunities. Readiness should be demonstrated through evidence of professional decision-making and instructional practice rather than compliance or simple counts of AI use.

This perspective also highlights the role of schools, districts and policymakers. Teachers cannot be expected to navigate these challenges independently. Professional learning, supportive organizational conditions and coherent policies all play a role in helping educators develop the expertise needed to use AI responsibly and effectively.

The goal is not to create teachers who use AI all the time.

The goal is to create teachers who know why they are using it, what they hope to achieve and when alternative approaches better serve students.