Reshaping Assessment: How GenAI is Transforming K-12 Evaluation and Feedback
Top trends and insights from The Edtech Insiders Generative AI Map research process about how Generative AI is transforming Instructional Materials
Reshaping Assessment: How GenAI is Transforming K-12 Evaluation and Feedback
By Alex Sarlin
As part of our updates to The Edtech Insiders Generative AI Map, we’re excited to release a new mini market map and article deep dive on Generative AI tools that are revolutionizing K-12 Assessment and Feedback (one of the most important aspects of education, both in terms of pedagogical efficacy and classroom logistics).
In our database, the Assessment and Feedback use case category encompasses tools that:
Improve and enable detailed and immediate feedback
Expand the scope of assessment
Rapidly generate assessment materials
Our updated mini market map for Assessment and Feedback showcases over 60 companies addressing seven distinct use cases. It’s worth noting that general-purpose AI tools like ChatGPT, Claude, and Gemini also excel at generating assessments and feedback. Additionally, tools focused on AI-Enhanced Tutoring, Study Material Generation, and Teaching & Learning Suites frequently incorporate robust assessment features as well.
As we look across the use cases in this category and how AI tools are evolving in the education space, we believe that AI is uniquely suited to transform assessment, which is a particularly anachronistic aspect of many educational experiences. With AI’s transformative capabilities, assessment is evolving from static, human-centered snapshots to dynamic, continuous systems that blend human insight with automated feedback.
Gen AI Trends for Instructional Materials
While we can’t encapsulate everything that is included in the rapidly evolving landscape of Gen AI and Assessment and Feedback, here are some of the top trends and insights we compiled while completing our most up to date research for The Edtech Insiders Gen AI Map:
Expanding What Can Be Assessed
AI is revolutionizing assessment by broadening what can be measured and how learning is demonstrated. Traditional tests often limit students to written responses or multiple-choice questions, but AI-powered tools enable a far richer evaluation landscape. Students can now express their understanding through multimodal projects—combining video, audio, images, and text—and across multiple languages. This approach not only assesses final products but also captures the learning process itself, making previously hidden aspects of understanding visible and measurable.
Imagine a high school assessment. Most would picture bubble sheets, written essays, or standardized digital exams. But with AI, new possibilities emerge: students singing a song to explain scientific concepts, recording a screen-share video with commentary to demonstrate problem-solving strategies, or collaborating on real-world projects are becoming within reach. You could even pitch ideas to a simulated Shark Tank panel or engage in week-long immersive simulations of college or workplace environments. AI provides real-time, scalable feedback on these tasks, aligning assessments with human-created rubrics while capturing complex problem-solving processes.
Tools like Goblins, Snorkl and Equatio provide rapid feedback in response to student handwriting and voice.
While some of these concepts sound fantastical, these ideas aren’t new. The Universal Design for Learning (UDL) framework has long recommended ‘multiple means of action and expression (that) encourage students to demonstrate their learning through various forms’. Advocates of Project-based learning (PBL) know the value of authentic, student-led projects and public portfolios. Education researchers have recommended “stealth assessment”, “choice-based assessment” and “curriculum-embedded assessment” for decades. And to be clear, thousands of sophisticated educators, schools (e.g., High Tech High), and pioneering edtechs (e.g., DIY.org) have provided… but only within small pockets.
Suddenly alternative, multimodal assessments are available at scale. Do you hear a bubble popping?
Classroom Example: In Ms. Rodriguez’s 8th-grade math class, students no longer submit paper worksheets. Instead, they record videos explaining their thought processes while solving geometry problems. Jamie, a student who struggles with written explanations but excels verbally, receives immediate, tailored AI feedback. The system identifies conceptual strengths, highlights minor calculation errors, and synthesizes common misconceptions across the class, allowing Ms. Rodriguez to adjust instruction and monitor each learner’s growth over time.
Accelerating the Feedback Loop
AI is transforming feedback from a delayed, static process into an immediate, dynamic engine for learning. Research consistently shows that timely, targeted feedback is one of the most powerful drivers of student achievement. Yet, traditional feedback methods—often returned days or weeks later—rarely meet the criteria for what decades of research deem effective: immediate, corrective, elaborated, actionable, and growth-oriented.
The greats in education have highlighted this for years:
How Learning Works: “Goal-directed practice coupled with targeted feedback enhances the quality of students’ learning.”
John Hattie: “Feedback is one of the most critical influences on student learning.”
Roediger and Butler: “The need for feedback is critical after any type of test”.
K. Anders Ericsson: “Deliberate Practice… involves the provision of immediate feedback, time for problem-solving and evaluation, and opportunities for repeated performance to refine behavior.”
AI-powered tools are changing this landscape. By delivering rapid, personalized feedback, these systems enable students to iterate continuously, refining their work through multiple cycles of improvement. Feedback becomes not just a final comment but a central part of the learning process. Importantly, AI can assess both the product and the process—providing real-time insights into students' reasoning, highlighting conceptual misunderstandings, and suggesting next steps, and not only leaving feedback to final products or assignments that have been “turned in.”
This TikTok testimonial and demo from an education tech coach for feedback tool Class Companion is useful for envisioning how AI feedback and grading looks today.
Tiktok failed to load.Enable 3rd party cookies or use another browser
While many will (and do) complain that AI-provided feedback cannot compare in quality to that of a trained professional educator, the other benefits of an always-available AI that can auto-grade against any framework will (and do) outweigh the drawbacks. Go to the website of any AI-Assisted Grading tool and you’ll also notice that the main value proposition for teachers is to ‘save time on grading’ - no small feat in an era in which “more than four in 10 K-12 workers in the U.S. (44%) say they are "always" or "very often" feel burned out at work, outpacing all other industries nationally (Gallup, 2022)”
Classroom Example: During a poetry writing assignment, Sarah uses an AI writing assistant to refine her work. Instead of waiting days for feedback, she receives immediate suggestions on metaphor usage and rhythm. Over five iterative drafts, Sarah integrates this feedback, deepening her understanding with each revision. By the time she submits her final version to Ms. Rodriguez, the teacher can concentrate on higher-order literary analysis, knowing foundational feedback has already been addressed by the AI system.
Infinite “Personalized” Practice Assessments
AI is redefining personalized learning by enabling limitless customized practice opportunities. While the term “personalized” is overused, AI makes true “precision learning” a reality. Learners and educators can now generate infinite tailored assessments, from quiz questions to complex case studies, aligned with individual goals, interests, languages, and learning levels—something previously impossible without extensive manual effort.
Before AI, edtech companies painstakingly expanded problem libraries through manual creation or basic software tools. Now, any educator or learner can instantly produce customized materials, adjusting content difficulty and feedback to meet evolving needs. The abundance of AI-powered quiz generation tools reflects this capability, though those that deeply understand pedagogy and seamlessly integrate with learning management systems and edtech incumbents like Kahoot!, Blooket or Quizlet are best positioned to lead (see testimonial for QuestionWell below).
Tiktok failed to load.Enable 3rd party cookies or use another browser
This capacity for infinite, adaptive practice doesn’t just support differentiated instruction; it empowers students to drive their own learning. By making practice relevant, engaging, and precisely targeted, AI fosters deeper mastery and sustained motivation.
Classroom Example: Ms. Rodriguez encourages her students to use AI tools to generate personalized practice sets at home. Miguel, working on algebra, receives dynamically adjusted problems based on his previous errors, while Aisha, who has mastered foundational concepts, tackles more complex challenges. The AI tracks their progress in real time, and Ms. Rodriguez reviews a dashboard that ensures each student’s practice aligns with classroom learning goals. This individualized approach not only deepens understanding but also keeps students engaged and challenged at the right level.
What’s Next: Building Comprehensive Learning Pictures
AI is poised to revolutionize assessment by offering a holistic view of student development, tracking growth across subjects and over time. By synthesizing diverse learning experiences, AI can create cohesive narratives of student progress, highlighting patterns in learning approaches and supporting a deeper understanding of individual capabilities.
One of the most promising aspects of AI-enhanced assessment is its ability to incorporate competencies and skills alongside traditional content knowledge. We’ve seen a recent embrace of the ‘portrait of a graduate’ movement in American K12 education in many different states, as well as a general interest in defining and measuring the competencies and skills that students need, which often take the form of a student ‘portfolio’, a ‘mastery transcript’ or both. However, the challenge lies in evaluating these complex attributes and translating them for use in admissions and career pathways—a challenge AI is uniquely equipped to address.
The need to define and evaluate ‘competencies’ through a combination of frameworks and student data, and to synthesize large amounts of learning data, both personal and aggregated, are both core capabilities of LLMs. At least, in theory. The amount of work it takes states or schools that embrace ‘portraits of a graduate’ to define, evaluate and report on the underlying competencies,‘skills’, or mindsets cannot be underestimated. Moreover, the resulting competencies are often limited to the school, district or (at best) state and will need additional translation (likely with AI) to be usable in admissions settings.
Classroom Example: At the semester's end, the AI system integrated into the school’s LMS analyzes David’s performance across math, science, and ELA. It identifies his strength in visual-spatial reasoning, which consistently supports his success in geometry and science labs. The AI also evaluates his ability to collaborate and think critically by examining classroom discussions and teacher feedback. This comprehensive portrait enables Ms. Rodriguez to tailor learning opportunities that align with David's strengths and preferences. Additionally, David receives a dynamic learning profile that he can use to showcase his capabilities for future educational and career opportunities.
Edtech Insiders is Sponsored By:
Calling All Education Entrepreneurs! The Milken-Penn GSE Education Business Plan Competition (EBPC) is accepting applications now through February 19, 2025. Compete for over $200,000 in cash and prizes and connect with global education leaders and investors. Since 2010, the competition has awarded $2 million, with winners and finalists securing over $200 million in funding.
The EBPC is the world’s premier competition for early-stage ventures tackling inequities in education through innovation. It's open to education ventures worldwide addressing learning at any level, with less than $2M raised or earned in the past year—no Penn affiliation required!
This edition of the Edtech Insiders Newsletter is sponsored by Tuck Advisors.
Since 2017, Tuck Advisors has been a trusted name in Education M&A. Run by serial entrepreneurs with over 25 years of experience in founding, investing in, and selling companies, we believe you deserve M&A advisors who work as hard as you do. Please enjoy access to The M&A Matrix by Tuck Advisors™ on ChatGPT—tailored to support founders and CEOs like you in navigating the M&A world.
Have questions, or are you ready to explore the next steps? Reach out to us.
Cooley LLP is the go-to law firm for edtech innovators, from early childhood through workforce. Informed by decades of experience in the education vertical, Cooley created the first edtech practice to provide industry-informed, business-minded counsel to companies and organizations at all stages of the corporate lifecycle. Cooley provides a multidisciplinary approach to client needs, offering seamless collaboration across offices and practices.
To learn more about what Cooley can do for you, reach out to Naomi May.
Edtech Insiders is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.