Page 83 - EW may 2026

Teacher-2-Teacher

What students should do

when AI prompts back

SUMANTH PRABHU

HERE IS A QUIET SHIFT HAPPENING IN India’s While students are learning how to ask
classrooms and homes. Children are not waiting
to be taught AI (artificial intelligence) as a subject. AI the right questions, few are being
TThey are already interacting with AI systems, of- taught how to interpret its responses,
ten without understanding how they work. From asking
chatbots to solve homework questions to generating im- challenge its output and understand its
ages, interaction with AI has become natural and almost limitations
instinctive. For most students, using AI is no longer a skill
to be learned but a behaviour they pick up through everyday process. Requesting examples or real-world applications
exposure. can provide greater clarity. For instance, asking for a step-
However, this rapid adoption arouses a deeper concern by-step explanation together with reasoning is far more ef-
that is often overlooked. While students are learning how fective than simply asking for a solution. Better prompts
to ask AI the right questions, few are being taught how to lead to more meaningful responses, but real learning still
interpret its responses, challenge its output, and understand depends on how students engage with responses.
its limitations. Knowing how to prompt AI is becoming com- Why AI needs structured learning. Students are al-
mon, but knowing how to respond after it responds, is rare. ready using AI independently, and this trend will acceler-
When AI starts prompting back. Modern AI systems ate. This is where structured learning becomes essential.
are no longer passive tools that respond to queries. They Schools need to move beyond grudgingly acknowledging
guide conversations, suggest next steps, and influence de- AI usage and begin guiding it. The transition from unstruc-
cision-making in subtle ways. In many cases, AI doesn’t just tured exposure to guided understanding ensures that stu-
answer questions, it shapes how next questions should be dents don’t just interact with AI but learn how to think while
framed. This has created a new type of interaction where AI engaging with it. Structured learning provides a framework
tools actively participate in the thinking process. where curiosity is supported by clarity and experimentation
For students, this introduces a critical moment in learn- is backed by comprehension.
ing. When AI suggests a method or provides a structured In this evolving landscape, platforms like Ulipsu play an
solution, the student must decide whether to accept it, ques- important role in bringing structure to AI learning. Instead
tion it, or explore alternatives. This decision determines of leaving students to explore AI tools without direction,
whether learning becomes deeper or remains superficial. Ulipsu integrates machine learning concepts, data literacy,
For example, when a student receives a step-by-step and ethical awareness into well-defined curriculums across
math solution, the real learning begins after the answer. different grade levels.
Understanding why the steps were taken, identifying pos- Learning AI through experience. Hands-on learning
sible errors, and applying the same logic to a new problem, plays a crucial role in developing understanding in the new
defines meaningful learning in the AI-driven era. AI age. When students engage with projects that involve
What should students do when AI responds? This training models and data analysis, they will understand how
directive is missing in most learning environments. Once AI AI systems function. They will understand how input data
provides an answer, students need to actively engage with shapes results and why output can vary.
it instead of passively accepting it. The first step is to verify The role of ethical awareness. As AI becomes more
whether the answer makes sense by cross-checking with integrated into everyday life, ethical awareness becomes an
reliable sources such as textbooks and teachers. This helps essential part of learning. Students must recognise that AI
students understand that AI output is not always accurate systems can reflect biases inherent in the data they sweep
and needs validation. up. Students also need to understand the dangers of using
The next step is to analyse the reasoning behind the an- AI in decision-making, especially in politically and socially
swer. Students should ask why a particular methodology is sensitive issues. Understanding who might be affected by an
used and whether alternatives exist. This encourages deep AI-generated decision and whether the output is unbiased
thinking and prevents blind dependence on AI-generated and accurate helps build responsible thinking.
responses. After that, applying the answer in different con- The future: Classrooms that teach thinking, not just
texts becomes important. If concepts are grasped, applying tools. The future of education will not be defined by access
them to new problems is the valuable outcome. Finally, as- to AI, as that is already becoming widespread. Instead, it
sessing what was learned and what remains unclear, helps will be defined by how well students can develop cognitive
reinforce learning and builds long-term clarity. skills in an environment where AI is omnipresent. Teach-
A practical note on better prompting. In the age of AI, ers will need to encourage students to pause, reflect, and
the teacher’s role is to teach students to ask for explanations question rather than tamely accept AI generated answers.
rather than answers. Breaking broad questions into sev- (Sumanth Prabhu is the Bengaluru-based CEO of edtech start-up
eral smaller parts makes concepts-learning a step by step Ulipsu)

MAY 2026 EDUCATIONWORLD 83