AI could cause patients harm as a result of bias and errors, health experts say

• Public health experts say AI poses a risk to human health, per the BMJ Global Health journal.
• They said AI has the potential to produce negative health impacts.

Public health experts are raising the alarm about the potential risks that artificial intelligence could have on human health.

Writing in the journal BMJ Global Health, health professionals from the UK, US, Australia, Costa Rica, and Malaysia said that while healthcare could benefit from AI, the technology "also has the potential to produce negative health impacts."

They said the use of AI in healthcare could "worsen social and health inequalities," either by incorporating human biases and discrimination or by "deploying AI in ways that reinforce social inequalities in access to healthcare."

They wrote: "Most of the health literature on AI is biased towards its potential benefits, and discussions about its potential harms tend to be focused on the misapplication of AI in clinical settings."

The experts also raised concerns about data privacy and security, as well as AI systems' tendencies to make errors that could "cause patient harm." The authors called for effective regulation around AI and for development to halt until it is implemented. They cited additional concerns around existential risks to humanity.

The warning is the latest in an explosion of concern around AI safety. AI experts, including one researcher dubbed the "Godfather of AI," have recently raised concerns against the new technology. A string of experts have been calling for the powerful tech to be regulated and have spoken about the potential dangers of an AI arms race between tech companies.

In March, Elon Musk and several AI heavyweights signed an open letter calling for a six-month pause on advanced AI development. The letter cited potential risks to humanity and society, including the spread of misinformation and widespread automation of jobs.