As artificial intelligence (AI) continues to permeate the realm of education, the largest provider of teacher licensing examinations is now offering a new test designed to assess educators’ proficiency in utilizing AI technologies. This move underscores the growing importance of AI in K-12 education.
With the rapid advancement and adoption of AI over the last three years, teachers find themselves at the forefront of this technological wave. Nevertheless, they currently face significant gaps in training and resources regarding the ethical and effective integration of AI into their teaching practices. The recently launched Futurenav Adapt AI test by ETS aims to clarify the competencies educators need in this evolving landscape.
This assessment evaluates educators through three scenario-based tasks, focusing on their ability to recognize and comprehend generative and large-language model AI within an educational framework. Additionally, it assesses their ethical navigation of the technology, evaluation of AI tools, and classroom application. The test typically takes less than 30 minutes to complete.
Currently, 46 states rely on ETS’s suite of Praxis tests to measure teaching competencies and subject-specific knowledge for teacher certification. While the new AI test wasn’t specifically designed for licensure or high-stakes purposes, it strategically positions ETS to accommodate any future requirements for AI-related topics in teacher certification.
“There’s this universal concern around appropriate use: for protecting students, ensuring they use it responsibly, helping teachers guide them, and making sure teachers are also using it correctly,” stated Vince Dean, associate vice president for the Praxis program.
Majority of Teachers Utilizing AI Yet Lacking Effective Guidance
According to the Center for Democracy and Technology, approximately 80% of teachers report incorporating AI into their classrooms—utilizing it for tasks such as creating lesson plans, developing scoring guides, designing quizzes, and detecting instances of AI-assisted cheating.
However, preliminary research reveals that AI-driven tools often deliver less substantive educational experiences compared to human instruction. Many educators struggle to formulate effective prompts to refine AI-generated content, and in some cases, these AI simulations can provide unrealistic depictions of students with varying ability levels.
While nearly all school districts have begun using AI-based tools or platforms, most teachers express that they are largely left to navigate these technologies independently. Despite some reports of training opportunities, many educators feel unprepared to effectively incorporate AI into their instruction.
“There’s a lot of choose-your-own-adventure stuff going on,” Dean remarked. “Engagement with diverse chatbots or platforms often leads to a lack of shared understanding of fundamental terminology: What is an LLM? How does it function? What considerations should be kept in mind when using them?”
For instance, test participants are required to determine how to utilize both public and private student data within an AI-integrated data tool. One scenario requires candidates to collaborate with a chatbot to design and execute a lesson plan, evaluating their proficiency in crafting AI prompts. The test’s concluding section involves participants reflecting on their experiences while working with a virtual instructional coach, also represented by AI.
This idea is not as far-fetched as it may initially appear: several districts have already begun employing AI tools within instructional coaching to analyze student data and offer feedback, as well as support teachers in reflecting on their pedagogical practices. The final portion of the AI assessment assesses educators not only on their skills but also on their feelings about working with AI technologies.
However, according to Akhil Verghese, founder of Krazimo, which aids districts in implementing AI systems, testing and professional development should not replace necessary guidance and support from school and district leadership. Currently, only two states, Ohio and Tennessee, mandate that districts establish an AI policy.
“Teachers are incredibly motivated and often take the initiative to find [AI] tools to integrate into their classrooms,” noted Verghese, who was not involved in the development of the AI skills assessment. “But even the most talented educators shouldn’t have to bear the sole responsibility for how AI is used in schools.”
Data from the AI test is made available through dashboards that highlight individual and group strengths, weaknesses, and potential equity concerns. Dean explained that administrators can leverage these dashboards to pinpoint areas for professional development or coaching and to plan AI implementation timelines.
ETS piloted the assessment last year with 75 secondary teachers from diverse fields, including English/language arts, math, history, and English as a Second Language, encompassing varying levels of teaching experience and familiarity with AI, according to Kateryna Komarova, a spokesperson for the organization.
