Ara Darzi says AI could speed up tests for drug-resistant infections

In short: Doctors often wait days to learn which antibiotics will work, and Ara Darzi says AI tools could speed up those decisions, but business incentives may slow real-world impact.

What's going on

Antibiotic resistance is rising, and it is already linked to more than 1 million deaths a year worldwide. It also contributes to nearly 5 million more deaths, according to figures cited by British surgeon Ara Darzi at WIRED Health in London.

One reason is that antibiotics are often overused or used when they are not needed. Bacteria can adapt when they survive exposure to antibiotics, which makes future infections harder to treat. A 2024 report in The Lancet estimated drug-resistant infections could cause 40 million deaths by 2050.

Darzi said a big problem is speed. Traditional lab tests often take two to three days because labs must grow bacteria from a patient sample. For sepsis, a life-threatening infection, doctors may not have that time, and each hour of delayed treatment can raise the risk of death by 4 to 9 percent.

Darzi argued that AI-powered diagnostics could help by quickly spotting resistance and guiding doctors toward the right drug sooner. He said some AI diagnostics are reaching over 99 percent accuracy without requiring new lab equipment, which could matter most in rural or remote areas.

What to watch

AI may also help researchers find new antibiotics and track how resistance spreads. But Darzi said many big drug companies have pulled back because antibiotics are not very profitable, since new ones must be saved for the sickest patients. The UK is testing a “Netflix-style” subscription, where the government pays a fixed yearly fee for access to new antibiotics instead of paying per pill, and Sweden is trying similar ideas.

Source: Wired