SINGAPORE – Imagine having a professor in your pocket, a research assistant working behind the scenes, or a debate partner waiting at your fingertips.
Chatbots and other artificial intelligence (AI) tools are fast becoming part of campus life, guiding and quizzing students through lessons.
As the AI boom hits universities, students are experimenting with technology to sharpen skills, get instant feedback and personalise their learning.
Students told The Straits Times that while AI does make learning feel too convenient, they check their work and ensure they still do the “primary thinking” themselves.
“It has the risk of making learners lazy, but when used properly, it can make the learning process very fast,” said National University of Singapore (NUS) law student Serene Cheong.
Dr Ranganath Vallabhajosyula from Nanyang Technological University (NTU) acknowledged that finding a balance between overusing AI and genuine learning is “tricky”.
“One strategy is to rely on AI for a basic level of learning, and then gradually… apply it to learn more advanced concepts,” he said, adding that the onus is on students to still take ownership of their learning.
“We should use AI more ergonomically, wisely,” Dr Ranganath said, noting that it is important to discern where AI adds value and plan its use so that it supports -– rather than replaces – critical thinking.
Law students at NUS use a chatbot to practise cross-examinations in a trial advocacy course.
Piloted in March, the bot simulates scenarios – from a distressed or evasive witness to a stern judge who demands precision – letting students practise questioning and courtroom etiquette.
After reviewing case documents, they act as defence counsel questioning their ‘client’, while a virtual judge and prosecutor respond in real time.
Year 3 law students Emmanuel Wong, 23, and Ms Cheong, 21, said the chatbot keeps them on track. Mr Wong noted that the ‘judge’ overrules them and corrects them when they ask leading questions – something their peers may not yet be trained to spot.
Ms Cheong said it reduces “psychological pressure” as she can practise at her own pace and regain composure.
Mr Wong added that the bot helped him ask precise questions and handle non-compliant witnesses or unreasonable objections.
While in-person training remains essential, the bot allows more personalised practice, testing students’ grasp of concepts before tutorials and serves as a gauge of their understanding, said Associate Professor Mervyn Cheong, who drew on his experience with real cases to develop hypothetical facts and scenarios for the chatbot.
Some 126 students have used the tool.
“If they plunge straight into in-person practices, when a lot of eyes are on them, it’s difficult to overcome the anxiety,” he said. “But if they do it in their own space, they have time to think and respond accordingly.”
Over at Singapore Management University, a design thinking bot plays the role of a stakeholder, as students negotiate and interact with clients. Rolled out in Jan 2024, nearly 400 students from the School of Computing and Information Systems have used it so far.
Koufu cleaner Ah Wu (centre) with SMU students (from left) Kayla Anne Vinh, Emma Foo, Ong Swee Long and Wong Jun Wei taking a photo with their prototype cart. The cart was designed to help track food waste in a way that fit the cleaner’s needs best.
PHOTO: COURTESY OF ONG SWEE LONG
Year 4 computing student Ong Swee Long, 25, used the bot in a sustainability course, where his group was tasked to propose a solution to tackle food waste on campus.
Initially drawn to a tech-driven approach, he said the AI prompted him to slow down and consider different perspectives.
For instance, as the bot walks students through the steps of design thinking, they are encouraged to engage with ‘stakeholders’.
For Mr Ong, this included students, the school’s provost, food stall owners and cleaners.
Conversations with the virtual “stakeholders” even prompted the group to speak to a Koufu cleaner in person, revealing practical constraints and perspectives a purely tech solution would overlook.
“It doesn’t replace talking to actual people, but it helps us to prepare some questions that we want to ask beforehand,” Mr Ong said.
The bot can generate prompts, problem statements, prototype options and even ideas for students to use or adapt.
SMU students practise negotiation, reasoning, and client interaction with AI. The bot walks students through four steps of the design thinking process – empathise, define, ideate, and prototype. During the first stage, students are encouraged to engage with relevant ‘stakeholders’.
PHOTO: SINGAPORE MANAGEMENT UNIVERSITY
Senior lecturer of information systems Kiruthika Ramanatha, who helped develop the tool, said its goal is for computer science students to think beyond technical fixes and consider human needs.
The AI-guided exploration ultimately led the team to design a centralised push cart with a weighing scale, cameras, and buckets to consolidate food waste tracking – replacing 16 carts the cleaner used to collect food from different stalls.
“Other than saving time and putting everything in one place, the AI encourages students to go through the whole process before jumping straight to a solution,” Mr Ong said.
Students are also treating AI like a round-the-clock study companion.
Several at the Singapore University of Technology and Design (SUTD) have taken learning even further by building their own bots to suit their learning styles.
Engineering Product Development undergraduate Anieyrudh R created “GPTBernie”, an AI tutor modelled on his professor Bernard Ee’s teaching style, the day before his mathematics final.
“Within minutes, it feels like I am in his office,” said the 22-year-old who cut his e-mails to his professor by 70 per cent. It was rolled out to a small group of students in April.
The bot helps him sort out more basic questions on his own, allowing face-to-face time with his professor to be spent on deeper, more complex conversations.
It also poses questions that challenge his understanding, said Mr Anieyrudh.
Engineering Product Development undergraduate Anieyrudh R created “GPTBernie,” an AI tutor modelled on his mathematics professor – complete with the professor’s teaching style and explanations.
ST PHOTO: LIM YAOHUI
He is leading teams of students who have, out of their own interest, developed similar bots to share with each other, expanding GPTBernie into a library of ‘GPTProfs’ that assist with administrative tasks, project design and teaching support.
At Nanyang Technological University’s Lee Kong Chian School of Medicine, AnatBuddy – an AI anatomy chatbot – helps first- and second-year students apply clinical knowledge.
Learning anatomy in a traditional setting is usually monotonous, said Dr Ranganath Vallabhajosyula, who helped develop the chatbot. AnatBuddy, which was introduced in November, changes that by pushing students to think deeper.
Instead of simply naming parts, students are prompted to explore clinical “what if” questions.
“They start thinking, ‘Where is it? What does it do? What happens if it’s damaged?’” said Dr Ranganath, adding that this reflects deeper engagement and stronger conceptual understanding.
Students also arrive better prepared and able to ask professors more substantive questions, he said.
For students at the Singapore University of Social Sciences, they have a full-on learning companion called the the iSmartGuide that dishes out bite-sized study units, powered by features like an AI tutor, and flashcard and quiz generator, so students can learn at their own pace anytime and anywhere.
SINGAPORE UNIVERSITY OF SOCIAL SCIENCES
Those at the Singapore University of Social Sciences have an AI learning companion called the iSmartGuide that dishes out bite-sized lessons, flashcards, and quizzes.
Since its launch in June 2025, the platform has served over 22,000 students.
Singapore Institute of Technology’s in-house platform, ClassAId, lets instructors create custom bots for any subject, from AI tutors to roleplay partners. As students interact with these bots, instructors gain real-time insights into their learning, enabling more targeted feedback.
Another SIT innovation, CommunicAId, acts like a personal communication coach that helps students rehearse professional scenarios, navigate tricky conversations or get instant feedback on assignments.
Research is another area that AI has proven to be useful in.
Since January, medical students at NTU have been using Rileybot, which helps them gather information from databases.
Senior medical librarian Rebecca David said that students are often confused or intimidated by traditional databases because each has a different interface.
NTU’s LKC Medicine has developed two bots for their students, one that trains them on anatomy, and another that acts a smart guide, helping students with scientific research.
PHOTO: NANYANG TECHNOLOGICAL UNIVERSITY
RileyBot guides them through accurate literature searches and is more precise than regular bots like ChatGPT, she said, adding that knowing how to carry out error-free searches is an essential skill for high-level research.
Over 500 students have used Rileybot and Anatbuddy so far.
“One bot is about technology, and the other is regarding medical anatomy,” said Dr Ranganath.
“When we bring these two together and bring it into the regular life of learning, it gives students more confidence and ownership of their learning.”