How AI Can Help Fight Antibiotics

Antibiotic resistance is of rapid growth people’s lives complications, which cause more than a million deaths worldwide each year and cause nearly 5 million more. These infections are more difficult and expensive to treat than more common infections, and result in longer hospital stays, raising costs for hospitals and patients alike.

Treatment often comes from a doctor’s guess. Ara Darzi, a surgeon and director of the Institute of Global Health Innovation at Imperial College London, says AI-powered infectious diseases offer a better way.

“We are standing, right now, in 2026, at the very beginning of this crisis,” Darzi said on April 16 at WIRED Health in London.

Misuse and misuse antibiotics and the lack of new drugs has been causing resistant insects. When bacteria are exposed to antibiotics that do not kill them immediately, they develop defense mechanisms to survive. Unwanted transcripts allow bacteria to develop immunity, rendering life-saving drugs ineffective. It means a reduction in the number of treatment options for patients with chronic diseases.

The problem is getting worse. A 2024 report in The Lancet He predicted that drug-resistant infections could kill 40 million people by 2050.

Traditional tests for antibiotic-resistant infections usually take two to three days, because they need to grow bacteria from samples. But some diseases, such as sepsis, are a time when patients do not have it. For every hour of delayed treatment, the risk of death increases between 4 and 9 percent. While waiting for the test results, doctors must use their judgment in choosing the drug to use.

AI-based insights can help inform those decisions. “AI-generated research is achieving more than 99 percent accuracy without the addition of a laboratory,” said Darzi.

These types of rapid information are especially important in rural and remote areas of the world, he added. The World Health Organization comparison that resistance to antibiotics is highest in Southeast Asia and the eastern Mediterranean, where one in three people said they were resistant to antibiotics in 2023. In Africa, one in five patients was infected.

AI can also help find new drugs for resistant diseases and predict the spread of resistant bacteria. The UK’s National Health Service is working with Google DeepMind to develop an AI solution to fight antibiotic resistance. In one demonstration, the system identified previously unknown pathways only 48 hoursunraveling a mystery that took researchers at Imperial College London a decade to understand.

Combined with an automated laboratory, Darzi said it is now possible to conduct hundreds of parallel tests over a period of time. Deep learning models can now analyze billions of cells in days, while artificial AI is being used to create things that don’t exist in nature.

However, major pharmaceutical companies have stopped developing antibiotics due to financial difficulties. New antibiotics must be developed to prevent resistance, but pharmaceutical companies benefit from high sales. There is little incentive for companies to stay in the game.

Darzi said new payment methods are needed to encourage the development of new antibiotics. In 2024, the UK launched a Netflix-style pay-as-you-go program in which the government pays an annual subscription fee to pharmaceutical companies to find new antibiotics, not the quantity. Sweden is also testing a partially bypassed version.

“The question that will determine the nature of medicine for the next 100 years, is not that we have the tools to answer. We have the tools,” he said. “The question is whether we have the character to think seriously about what we see.”