Assessment Evolved: Redefining formative assessment in a GenAI era

Assessment Evolved: Redefining formative assessment in a GenAI era

The rapid rise of Generative AI (GenAI) isn’t just introducing new tools. Globally, it's reshaping how students engage, how teachers teach, and how we measure learning.

While this transformation brings excitement and opportunity, it also raises urgent questions:  

  • How do we keep assessment meaningful when technology can offer instant answers?  
  • How do we protect learning integrity while preparing students for a future where AI is everywhere? 

Formative assessment has always been the cornerstone of effective teaching. It’s how educators check in with students, adapt their approach, and help them grow. In the age of GenAI, formative assessment sits at a crossroads. Used thoughtfully, AI can deepen reflection and feedback, preparing students for the future. Used uncritically, it risks compounding learning gaps and undermining the development of essential skills. If students rely on AI for quick answers, they miss out on the critical thinking and creativity that real learning requires. This can reduce opportunities for meaningful engagement, as students may feel that AI can do the work for them.

And the stakes are high. Most educators - 80% in fact – say formative assessment matters deeply and use it often. But just as many are worried about how students might use AI for assignments when prohibited. It’s a clear signal that educators need strong leadership and practical guidance. 

Signals that can't be ignored

Our recent survey of over 1,000 teachers and instructors in the UK and US reveals both concern and commitment.  

68% of schoolteachers and 82% of higher education instructors are worried about students using GenAI on formative assessments where it’s not allowed. It's not just about cheating—it’s about the risk of students missing out on the skills and knowledge they’ll need for the future. 

“The rhetoric of ‘AI is for cheating’ is so loud that the greater ethical concerns underneath are either being ignored, or not acknowledged… The deeper reason why academic integrity breaches are so scary right now is the fear that kids will start devaluing the role of education, the role of learning.” — Amanda Bickerstaff, CEO, AI for Education 

But educators aren’t shying away. Many are already experimenting with GenAI in their classrooms, looking for ways to use it responsibly and effectively. For example, around half of educators are integrating AI tools into learning activities and assessment design. And GenAI use has taken off among university students in the UK—jumping from 53% in 2024 to 88% in 2025.  But the question shouldn’t be “AI: good or bad?” Instead, it’s about helping students use AI thoughtfully, keeping learning at the center. 

Can AI detection solve these issues? The short answer is no 

Much of the conversation around AI in education has focused on detection and restricting use. It’s tempting to rely on things like AI detectors to catch AI misuse, but research shows these tools aren’t reliable, and they can erode trust between teachers and students. As Rachel Evans, Director of Digital Transformation at the Girls’ Day School Trust, notes:  

“We’ve moved from educators feeling very confident 18 months ago that they would be able to identify AI writing in their students’ work—to now, where it’s really much more difficult to detect.”  

The solution?  Designing assessments that make responsible AI use part of the process. When educators are confident in using GenAI themselves, they can model best practices and help students build the habits of questioning, critiquing, and connecting ideas. 

Trust in learning comes from purposeful, transparent assessment, not from detection or bans. When AI is integrated intentionally, it can build integrity, deepen learning, and strengthen teacher-student collaboration. 

Building AI literacy and future-ready skills  

AI literacy isn’t just about knowing how to use the tools. It’s about knowing when to use them (or not), trust them, and how to build on what they offer. Effective formative assessment in the AI era means helping students:  

  • Question outputs, not just accept them
  • Connect ideas and seek alternative perspectives 
  • Build habits of reflection, critique, and deeper understanding 

Thoughtful formative assessment can turn GenAI into a true learning partner—balancing AI interactions with students’ own judgment and experience. These future-ready skills, along with the ability to summarize, ideate, and create with or without AI, will be essential for tomorrow’s workplace. 

As Ryan Baker, Professor of Artificial Intelligence and Education, puts it: “The more that we can be in the space of allowing people to use the tools, helping them learn how to use AI effectively, and then holding them responsible for the content - I think it’s a much better world to be in than trying to prevent use…These tools are a reality. Why would we train somebody not to use the tools that they’re going to use for the world of work?” 

Practical strategies for evolving assessment

So, how do we design assessments that reveal real learning? Our report offers four key questions for educators to ask when designing tasks in the GenAI era:  

  • What do I want to assess, and does the task show evidence of that learning?
  • Does completing this task require the skills I want students to develop?  
  • Will misuse of GenAI compromise its learning value?  
  • Could thoughtful use of GenAI enhance its learning value? 

By reflecting on these questions, educators can redesign formative assessment activities to gather richer evidence of student understanding, and to build the skills that matter most. 

 Change can’t happen in isolation 

The best solutions come from partnering with educators. Teachers know their students, their subjects, their communities. But bold action is needed across all levels of education for proactive, future-ready formative assessments. School and higher education leaders, policymakers, and administrators have a vital role to play:  

  • Policy: Partnering with teachers to design and review AI integration initiatives  
  • Training: Giving educators ongoing opportunities to build confidence and skill with GenAI tools 
  • Broader ecosystem: Facilitating dialogue about GenAI’s impact on curricula, instruction, and assessment 

Federal, State, and local governments implementing AI strategies must work in partnership with all stakeholders, providing regular opportunities for review, as well as ensuring educators have access to ongoing training and support. It’s also time to work together to better teach and assess AI literacy, improve alignment of academic standards and summative assessments with the skills students will need for the future. When students develop these skills, they’re better prepared to thrive in tomorrow’s workplaces and beyond.

The future won’t wait—neither should we 

GenAI is here to stay, and its impact will only grow. The challenge is real, but so is the opportunity. This is just the beginning of an important conversation on how formative assessment can adapt to this new era. By evolving assessment practices, building AI literacy, and supporting educators, we can prepare students for a world that keeps learning at the center.  

Ready to dive deeper? Read Assessment Evolved in full for practical strategies, expert insights, and real-world examples.

completely agree with this. At the same time, there is much deeper complexity behind how these models actually work. As foundation models continue to evolve, the pace of technological change is creating new challenges that educators, learners, and institutions are only beginning to face. While we now have powerful capabilities, the ethical and responsible use of these models remains a significant challenge. Understanding not just what these systems produce, but how and why they produce it, will be critical in the next era of learning and assessment.

Congratulations on the first issue — what a thoughtful way to explore how AI and education can evolve together. I appreciate the emphasis on curiosity, responsibility, and access, especially in a space that shapes so many learners’ futures. Excited to see where this perspective continues to grow at Pearson. 🌱

When formative assessment is rethought with GenAI in mind, it shifts from monitoring performance to supporting deeper learning. Done well, it can strengthen academic integrity while helping students build the judgment and skills they’ll need to work responsibly with AI.

This is, of course, timely and positive. Such positivity is the right approach, while so many of us are feeling overwhelmed by the pace and diversity of #GenAI. So we need to act fast. My students have fallen on AI like a swarm of locusts in a feeding frenzy. They are already massively dependent on, for example, Chat GPT. They are not going to give it up without a fight. If we are to maintain academic integrity, on both sides of the teaching/learning context, we have to find ways of safeguarding critical thinking skills, while training students to exploit AI with genuine transparency and accountability. Currently, in spite of all the advice and wisdom in this Pearson Perspective, I’m very unsure how we can achieve this in realistic ways. It’s a bit like giving students a high performance automobile, but cautioning them to keep the speed down to 10 mph…

Dear global friends, wish you a very happy new year 2026 for the new beginnings of long waited peace and happiness , let’s forgive the people who had given deadly betrayal pains , let them be happy even the pain remains with us since it is our sacrifices for theirs life, let us thank the people who were tried to heal us in the place of god, lets hope we can achieve the global goal of peace and happiness and non poverty and sustainability to every one in the World without any partiality., Vijayakumar Varadarajan

Like
Reply

To view or add a comment, sign in

More articles by Pearson

Others also viewed

Explore content categories