AI SCIENCE
The cure to superbugs might not come from humans
AI has come up with two experimental antibiotics that could fight drug-resistant gonorrhoea and MRSA, according to a team at MIT.
The drugs were designed atom-by-atom using generative AI and proved effective in both lab and animal tests.
But they’re still years away from being prescribed, as further refinement and clinical trials are needed.
Drug-resistant infections now kill more than a million people each year.
Overuse of antibiotics has helped bacteria adapt, while new drug discovery has slowed for decades.
Earlier AI studies scanned existing chemicals for potential antibiotics.
This research went further, teaching AI how bacteria respond to molecules so it could generate brand-new designs from scratch.
From 36 million options, two candidates advanced.
They worked against gonorrhoea and MRSA in mice, but scaling up is far from simple.
Many AI-designed compounds can’t actually be manufactured, and the drugs must also prove safe in people before moving forward.
Researchers say the findings could signal a “second golden age” for antibiotics, though challenges remain.
Here’s what you should know:
AI created two new antibiotic candidates effective in lab and animal testing.
Clinical trials are still years away.
Safety, manufacturing, and lack of commercial value remain major barriers.
New antibiotics are difficult to bring to market, not just for scientific reasons, but because they’re rarely profitable.
To preserve effectiveness, they need to be used sparingly, which limits commercial incentive.
Pharma execs: 😬