How Medical Interns Use the DeepSeek Chatbot for Training

In short: A study found that about one-third of medical interns at one hospital use the DeepSeek chatbot for tasks like school assignments and translating text. This evidence comes from a single questionnaire study with a specific group of students.

Artificial intelligence tools are moving into medical schools, but how are students actually using them behind closed doors?

What happened, in plain words

Researchers at the Second Affiliated Hospital of Anhui Medical University surveyed 634 medical interns to see how many use an artificial intelligence chatbot called DeepSeek. They found that roughly one-third of the students reported using the tool. Among those users, over half relied on it several times a week for tasks like finishing assignments, searching for facts, translating languages, and polishing written work. The study also showed that urban students and nursing interns were more likely to use it, while preventive medicine interns used it less than clinical medicine students.

Key points

One-third usage rate: About 634 interns completed an anonymous questionnaire, and roughly one-third of them reported using the DeepSeek chatbot for their day-to-day routines.
Regular writing and searching: Most students who used the tool did so several times a week to help with homework assignments, look up information, translate languages, and refine written text.
Mixed student backgrounds: Researchers noted that living in an urban area and studying nursing were linked to a higher likelihood of using the chatbot, while preventive medicine students used it less often.
Mostly neutral attitudes: Students generally felt neutral or optimistic about the technology, with most users believing its benefits and drawbacks were roughly balanced.

Terms explained

Cross-sectional study — A type of research that looks at a group of people at a single point in time to see what is happening. Example: Taking a snapshot of what people are eating for lunch today at a school cafeteria.
Large language model — An artificial intelligence system trained on massive amounts of text to understand and generate human-like writing. Example: A computer program that guesses the next word in a sentence to help draft an email.
Medical intern — A student or recent graduate training in a hospital to become a doctor or nurse. Example: A medical student shadowing physicians in a hospital ward to gain hands-on practice.

Critical analysis

What is encouraging. The study offers a helpful and structured look at how real students in a specific hospital setting integrate domestic artificial intelligence tools into their everyday training routines.

What deserves caution:

Single hospital survey. The research collected answers from only one institution, meaning the habits of these specific interns might not reflect medical students everywhere else.
Self-reported answers. The findings rely entirely on anonymous questionnaires filled out by students, which can sometimes lead to people misremembering or overstating their actual habits.
No quality check on output. The study tracked how often students used the chatbot and what they used it for, but it did not test whether the chatbot's answers were accurate or helpful for medical learning.

Bottom line: While about a third of surveyed interns use the DeepSeek chatbot, the study only shows basic usage patterns rather than proving whether the tool improves medical education.

What we still don't know

The source does not cover the accuracy of the information provided by the chatbot, nor does it track the long-term effects of using artificial intelligence on student test scores or clinical skills.

Source: PLOS ONE. The original is licensed CC BY 4.0. This text is an AI-assisted adaptation (summarized, simplified and translated) and may differ from the original.