Dar es Salaam. Scientists have cautioned that the growing trend of using artificial intelligence (AI) tools for self-diagnosis and treatment is delaying patients from receiving timely medical care and worsening drug resistance, prompting the government to issue guidelines.
The use of AI in self-medication, such as mental health apps and symptom-checking systems, has been rising.
However, experts warn that misuse or over-reliance on such tools can lead to serious health risks.
They noted that AI can sometimes give incorrect or incomplete diagnoses due to inaccurate data, lack of professional evaluation, or confusion between similar symptoms, such as those of malaria and influenza.
“This may delay access to proper care, as many people treat AI-generated responses as medical advice and skip hospital visits—causing diseases to worsen or become chronic,” said scientists attending the Tanzania Health Summit, which brought together 1,700 health professionals, researchers, pharmacists and policymakers from 25 African countries.
Speaking after the meeting, Tanzania Health Summit Board chairperson Chakou Tindwa said while technology has revolutionised healthcare delivery, its rapid use raises regulatory and ethical concerns.
“Scientists are asking: can a person input symptoms into AI and trust the response enough to buy medication? This area still needs strong supervision,” he said. “We must ensure accuracy, safety and oversight while addressing fears among doctors that AI could replace them.”
Dr Tindwa added that technological improvements must not compromise service quality or public safety.
“For instance, with telemedicine, patients can now consult doctors via video calls from home. A doctor can monitor a patient’s heart remotely, but that doesn’t replace physical examination.”
Government stance
Chief Medical Officer Grace Magembe told said that AI in healthcare is not inherently bad, Tanzania must embrace it as the world advances, but warned that its use requires caution.
“There are areas where AI can be helpful and others where it should not be used,” she said. “If you tell AI that you have a headache, fever and cough, it may suggest a condition but there are over 100 illnesses with similar symptoms.
Only a doctor can properly assess your environment, medical history and test results to determine the cause.”
Dr Magembe said AI cannot replace medical consultation. “You cannot tell AI what’s in your chest and expect a correct diagnosis. It’s useful, but it cannot remove the need for professional examination,” she said.
She acknowledged that AI performs well in assisting doctors with test interpretation, such as X-rays or CT scans, but only after an initial medical review.
“We welcome AI in healthcare and have started using it, but with precautions. Self-treatment contributes to drug resistance, which is already a major problem,” Dr Magembe said.
According to her, many patients misuse AI advice, buy the wrong medicines and delay diagnosis. “Technology is good, but we must approach it carefully.”
The summit also discussed three key themes: data use, technology and universal health coverage.
Experts called for the integration of public and private health data systems, allowing real-time tracking of patients and disease trends.
“It would be ideal if the Ministry of Health could access data from both public and private hospitals nationwide through one platform,” Dr Tindwa said.
On universal healthcare, scientists stressed the need for accessible, affordable and high-quality services for all citizens.
Register to begin your journey to our premium contentSubscribe for full access to premium content
