Professor Konstantina Nikita Discusses the Risks and Limitations of AI-Based Medical Self-Diagnosis in Kathimerini

Professor Konstantina Nikita was featured in the Greek newspaper Kathimerini in an article titled “ChatGPT, What’s Wrong with Me? – The Pitfalls of Digital Self-Diagnosis”.

In the article, Professor Nikita discusses the limitations and potential risks of using generative Artificial Intelligence (AI) tools for medical advice and self-diagnosis, highlighting the importance of clinical validation, human oversight, and appropriate safety mechanisms.

She also presents ongoing research conducted by the BIOSIM team in collaboration with researchers from Harvard University and the National and Kapodistrian University of Athens, evaluating the diagnostic performance of large language models (LLMs) in complex medical cases.

The full article is available here.