Khan Academy is expanding its AI tools to change how schools measure learning. New software like "Explain Your Thinking" and "Khanductor" aim to focus classrooms on student reasoning instead of final answers. This transition comes as educators nationwide struggle to assess student understanding when generative AI can instantly produce polished homework.
What Happened
According to school administrator Dr. Tim Krieg in a post on the Khan Academy Blog, schools must redefine what counts as learning in an AI-shaped world. Krieg argues that policy and data dashboards alone cannot save education. Instead, schools need to focus on how students think rather than just their finished work. Tools within the Khan Academy suite, including Khanmigo, Khanductor, and Writing Coach, are designed to make the work of student learning visible to teachers.
Rather than acting as simple answer generators, these applications guide students through the steps of problem-solving. As we previously reported, Khan Academy has focused heavily on reshaping its AI tutor to prevent students from skipping difficult steps. This aligns with an effort to ensure that technology is a "sit-beside" assistant that supports teachers, rather than a tool for compliance.
The Bigger Picture
This shift from grading final products to evaluating the learning process is backed by educational research. A study in Developing Academic Practice warns that grading only final homework assignments is no longer reliable because AI can construct polished artifacts that mask a student's actual skill level. Researchers writing for the Pacific Journal of Technology Enhanced Learning call this "cognitive offloading," noting that outsourcing intellectual tasks to AI leads to long-term declines in critical thinking and student ownership of learning.
Socratic AI tools can counteract this if designed properly. An independent study published on Zenodo tracked secondary school students using Khanmigo and found a statistically significant rise in mathematical competency and problem-solving skills. Khan Academy's internal testing also showed that prioritizing active student engagement over passive reading improved "next-item correctness," showing that students could solve subsequent problems independently.
However, software alone is not enough. A qualitative case study on Zenodo reveals that AI classroom tools fail to take root unless school districts provide clear policy guidelines and thorough teacher training. Similarly, in writing instruction, research published in Frontiers in Education shows that while AI offers rapid feedback on grammar and organization, human teachers remain indispensable for guiding complex arguments. Without this balance, students risk losing autonomy and creative thinking skills.
What This Means for Families
For parents, this means the traditional metrics of academic success are changing. A completed worksheet or a flawless essay draft is no longer proof that a child understands the material. Parents should expect schools to focus more on classroom discussions and multi-step drafts where students must defend their choices.
As we noted when discussing why Khan Academy changed its AI tutoring model, modern educational technology aims to make learning active rather than easy. Families should encourage children to view AI tools as partners in solving problems rather than shortcuts to a final grade.
What You Can Do
To support this shift at home, parents can ask children to explain the steps they took to reach a solution rather than just checking for correct answers. When children use AI homework assistants, parents should ensure they use Socratic settings that ask questions instead of tools that simply write answers. Finally, parents can advocate for local school boards to invest in professional development so teachers have the time to learn how to use these new tools effectively.