Artificial intelligence use, perceptions, and educational needs among medical undergraduates at a Sri Lankan medical faculty: A cross-sectional survey

The Digital Co-Pilot in the Doctor's Toolkit

Imagine the world of medicine as a massive, ever-expanding library. For centuries, doctors and students have relied on thick textbooks and their own memories to find the right book to solve a patient's puzzle. But recently, a new kind of librarian has arrived: Artificial Intelligence, or AI. Think of AI not as a robot that replaces the doctor, but as a super-fast, super-smart search engine that can read millions of medical books in a second and summarize the answers for you. It's like having a magical assistant that can help you write essays, explain complex diseases, or even suggest what might be wrong with a patient based on their symptoms.

This new assistant is changing how the world works, but it's also raising a big question for the future: How do we teach the next generation of doctors to use this tool? If a student uses a calculator to do math, they still need to understand the math to know if the calculator is right. Similarly, if a future doctor uses AI to help diagnose a patient, they need to know when to trust the AI and when to double-check it. This is the story of a recent study that looked at medical students in Sri Lanka to see how they are using this digital co-pilot, what they think about it, and whether their schools are teaching them how to use it safely.

The Study: A Snapshot of Student Habits

Researchers at the Sabaragamuwa University of Sri Lanka decided to take a snapshot of the medical student population to see what was happening in their digital world. They sent out an online survey to nearly 400 students, asking them about their habits, their confidence, and their feelings about AI. The results were like a peek into a classroom where almost everyone has a smartphone, but only a few have been taught how to use the most powerful apps on it.

The Digital Takeover
The first thing the study found was that AI has basically taken over the student body. Out of the 369 students who answered, a staggering 99.2% said they use AI tools for their studies. It's almost as if not using AI is the exception, not the rule. The most popular tool by far was ChatGPT, used by 98.1% of the students who use AI. It was followed by Google Gemini (34.9%), DeepSeek (20.2%), and Microsoft Copilot (6.2%).

When asked why they use these tools, the students were honest. The top reason was subject matter reference (89.3%), meaning they use it to look up facts and study for exams. The second most common use was general information lookup (72.9%). However, a significant portion—50.8%—admitted to using AI for assignment assistance, and 39.6% used it to improve their writing skills. This suggests that while many use AI as a tutor, a large number are also using it as a shortcut for their homework.

Confidence vs. Reality
Here is where things get interesting. The students were asked to rate how good they are at using these tools. For ChatGPT, 92.4% of the students rated their skills as "Good" or higher. They feel like they are masters of the tool. However, the study authors warn that feeling confident doesn't necessarily mean they are experts. Just because you can ask a question and get an answer doesn't mean you know if that answer is actually true or safe for a real patient. It's like feeling confident driving a car because you know how to press the gas pedal, but not knowing the rules of the road or how to handle a sudden storm.

The Medical Apps
Beyond the general chatbots, students also use AI-powered medical apps. The most popular was Medscape, used by 77.9% of respondents. This was followed by UpToDate (37.4%) and WebMD (15.4%). These are like specialized medical dictionaries that have been upgraded with smart features to help doctors find the latest treatments.

The Changing Attitude: From "Yes!" to "Wait..."
One of the most fascinating discoveries in the paper is how students' feelings change as they get older and move from classroom learning to real hospital training.

• The Freshmen and Sophomores (Years 1 & 2): These students, who are just starting their journey and learning the basics of biology and anatomy, were the most excited. 85.6% of Year 1 students and 75.6% of Year 2 students agreed that AI should be part of their medical curriculum. They see AI as a cool, helpful tool that can make learning easier.
• The Seniors (Years 4 & 5): As students move into the clinical years, where they start seeing real patients, their enthusiasm cools down a bit. Only 51.9% of Year 4 students and 61.4% of Year 5 students agreed that AI should be in the curriculum.

The researchers suggest this shift happens because older students are starting to realize the limits of AI. They are seeing that while AI is great for memorizing facts, it might not be as good at the messy, human side of medicine, like talking to a scared patient or making a tough decision when the data isn't clear. The study found that pre-clinical students were significantly more likely to believe AI could replace some doctor tasks, while clinical students were more skeptical.

The Ethical Worry
Despite the excitement, the students aren't blind to the risks. A large majority (63.4%) agreed that AI will bring new ethical challenges. They are worried about things like privacy, whether AI might make mistakes, and if relying on it too much could weaken the bond between a doctor and a patient. About 39.9% worried that AI might weaken the doctor-patient relationship, and 57.5% thought it would introduce new ethical problems in practice.

What This Means for the Future

The paper concludes that while Sri Lankan medical students are using AI more than ever, they are mostly doing it on their own, without much formal guidance from their schools. They are like explorers using a new map without a compass. The study suggests that medical schools need to step in and teach students how to use these tools responsibly.

The authors argue that we need to integrate AI education into the curriculum, not just to teach students how to use the tools, but to teach them how to think critically about the answers those tools give. They want to ensure that future doctors can use AI to become better physicians, rather than letting AI replace the human judgment that is so crucial in medicine. The study doesn't say AI is a magic solution or a dangerous monster; it simply says that the students are ready to use it, but they need a guide to make sure they use it safely and ethically.

In short, the future doctors are already talking to the robots, but they are waiting for their teachers to tell them what to say next.