Patterns of Generative Artificial Intelligence Use and Medical Students' Perceptions of Learning Among Eighth Semester Medical Students at Universidade Católica Timorense (UCT), Timor-Leste

Imagine the world of learning as a massive, sprawling library where every student is trying to find the right books to build a house of knowledge. For centuries, the only way to find those books was to walk the aisles, ask a librarian, or hope you picked the right shelf. But recently, a new kind of librarian has arrived: Generative Artificial Intelligence, or GenAI. Think of this not as a robot that knows everything, but as a super-fast, super-creative research assistant that can read a thousand books in a second and then write a summary, create a quiz, or explain a tricky concept in plain English. It's like having a study buddy who never sleeps and has read every textbook ever written, but with a catch: sometimes, this buddy gets so excited about telling a story that they might accidentally invent facts that sound real but aren't.

This new assistant is changing how students everywhere learn, especially in medicine, where the "books" are complex human bodies and the stakes are high. The big question isn't just "Can we use this tool?" but "How are students actually using it, and do they trust it?" A team of researchers wanted to peek behind the curtain to see how medical students in Timor-Leste are interacting with this digital helper. They weren't trying to prove that AI is magic or that it's dangerous; they just wanted to take a snapshot of reality to see if students are using it as a helpful sidekick or if they're getting lost in its chatter.

---

The Study: A Snapshot of Timor-Leste's Future Doctors

The researchers decided to hang out with a specific group of students: the 76 eighth-semester medical students at the Universidade Católica Timorense (UCT) in Timor-Leste. These are the students who have finished most of their classroom theory and are getting ready to dive into the real world of hospitals. The team asked them to fill out a survey about their habits with GenAI. They wanted to know: What tools are you using? How often? What are you asking it to do? And most importantly, do you trust it?

The Tools of the Trade
When the students opened their digital toolboxes, one name stood out like a giant among giants: ChatGPT. About 59.2% of the students (that's 45 out of 76) said they used it most often. But they weren't just using one tool; it was more like a toolbox with many gadgets. Google Gemini was the second favorite, used by 52.6% of the group, followed by Microsoft Copilot (42.1%) and Claude (36.8%). A few students also tried out DeepSeek, Perplexity AI, and other tools, but the big four dominated the scene.

How Often and Why?
The students weren't just peeking at these tools occasionally; they were using them like a daily vitamin. Nearly half of the group (46.1%, or 35 students) said they used GenAI every single day. Another 28.9% used it several times a week. It wasn't just for fun, either. They were using it for very practical, study-heavy tasks. The most popular job for the AI was summarizing PowerPoint presentations (89.5% of students), followed closely by making multiple-choice questions to test themselves (85.5%). They also used it to understand medical concepts (78.9%) and to prepare for exams (73.7%). It seems the students were treating GenAI like a personal tutor that could instantly turn a boring lecture slide into a clear, easy-to-digest story.

The Hiccups in the System
However, using this super-assistant wasn't always smooth sailing. The biggest headache wasn't that the AI was "stupid," but that it was often free and had limits. A massive 81.6% of students (62 of them) complained about hitting daily usage limits on free versions, which forced them to juggle multiple accounts just to keep studying. Internet access was the second biggest problem (72.4%), which makes sense if you imagine trying to talk to a robot while your Wi-Fi is acting up.

Then there were the "hallucinations." About 46.1% of students worried that the AI would make up facts that sounded real but were wrong. 55.3% were concerned about inaccurate information in general. It's like having a friend who tells a great story but occasionally mixes up the names of the characters. The students knew they had to be the editors, checking the facts before believing them.

What the Students Thought
Despite the glitches, the students' attitude was surprisingly positive. When asked if GenAI helped them understand medical concepts, 52.6% agreed and 32.9% strongly agreed. They felt it made learning more efficient and helped them prepare for exams. They also felt it was easy to use.

But here is the most important part: they didn't blindly trust it. When asked if they trusted the information generated by the AI, only 18.4% strongly agreed, while 40.8% just agreed. However, a huge majority (46.1% agreed, 38.2% strongly agreed) said they always verify the AI's answers using reliable medical sources. They treated the AI like a draft writer, not the final editor. They knew that before using the AI's advice in a real hospital or on a test, they had to double-check it against the "real" books.

The Takeaway
The study suggests that for these future doctors in Timor-Leste, GenAI has become a vital, everyday part of their learning routine. It's not replacing their teachers or textbooks; it's acting as a powerful supplement that helps them summarize, quiz, and understand complex ideas faster. But the students are smart enough to know the tool's limits. They are using it heavily, but they are also keeping a critical eye on it, making sure they don't get fooled by a "hallucination." The researchers believe this balance is key: using the technology to boost learning while remembering that the human brain is still the one that needs to do the final thinking and checking.