Antibiotics from AI? New Developments to Combat Antibiotic Resistance 

20th August 2025

By Mia Turner, Scientific Communications and PR Intern. 

Antibiotic resistance is responsible for around 5 million deaths per year due to the growing ability of bacteria to evade disarming and death by previously effective antibiotics. In recent years, scientists in the infectious diseases field have become increasingly concerned as more bacteria are developing resistance against antibiotics. As such, antibiotic resistance is rapidly becoming a critical global health issue in need of intervention because it “puts many of the gains of modern medicine at risk” by making infections more difficult to treat and common surgeries such as caesarean sections far riskier. However, a new study by a group of researchers at MIT was published in Cell last week, capturing the attention of experts and mainstream media alike, suggesting the potential development of new antibiotics and providing hope for a new antibiotic ‘golden age’. 

What is Antibiotic Resistance? 

Occurring when bacteria develop the ability to defend against antibiotics designed to kill them, antibiotic resistance has become a steadily growing issue. Once a type of bacteria becomes resistant against antibiotics it can continue to grow and spread, becoming incredibly difficult or potentially impossible to treat with any previously used antibiotic and a massive issue for global healthcare systems.  

This is the case with some strains of Staphylococcus Aureus, particularly MRSA (methicillin-resistant staphylococcus aureus) which can still fortunately be treated by different lines of treatment, and gonorrhoea for which there is only one remaining class of antibiotics that is effective, cephalosporins. However, even for the remaining applicable antibiotics, resistance is already showing signs of development which is a serious cause for concern with it being such a common infection. 

Antibiotics: The (Former) Golden Age 

The ‘golden age’ of antibiotics was an era of large biological development and discovery where many new and effective antibiotics were introduced into medicine, providing giant leaps in combatting bacterial infections. Penicillin, discovered in 1929, and the following discovery of streptomycin in 1943 marked the beginning of the age of antibiotics. The subsequent discovery of more than half of the antibiotics still used today between 1950 and 1960 resulted in large reductions in mortality rates from bacterial infections.  

However, since 1970 only eight new classes of antibiotic have been approved showing a severe slow-down in the pipeline. Pharmaceutical companies shifted their focus away from antibiotic development towards more profitable areas such as chronic disease treatment as it offers a far more reliable and long-term source of income as opposed to the shorter-term profits of antibiotics. As antibiotic use becomes more protected due to increasing risk of resistance, pharmaceutical companies show far less desire to develop new antibiotics for commercial use.  

New Antibiotics – with some help from AI 

With the lack of attention paid to the antibiotic pipeline in recent years, news last week of a new development with AI provided cause for excitement in the scientific community. Mainstream media, including the BBC, as well as trade press were quick to jump on the story out of a research lab at MIT that made the leap in using generative AI to design brand-new antibiotics, based on specific guidance and existing research. The study, published in Cell, centred on an AI platform that was given data on the chemical structure of known compounds with correlating data showing the effects each compound has on bacterial growth. The AI was then able to discern how bacteria are affected by different chemical compounds. 

Two different approaches were used in the generation process, one which gave the AI some moderate framework to start from and another which gave it complete freedom to design from the start.  Ultimately, the AI was able to create antibiotics for gonorrhoea and MRSA, which were then tested on infected mice in a lab and produced positive results. The process of drug development is a lengthy one – as we mentioned in last week’s article – but with the right resources it can become a more efficient process with current estimates hoping to progress these drugs to human clinical trials within the next few years.  

This new research shows promising steps towards finding new antibiotics and using a new method that hasn’t been exhausted by decades of existing research. The sheer volume of microbes developing resistance against formerly applicable antibiotics is a cause for concern for the scientific community but hopefully progress can be made to reduce the effects of antibiotic resistance on global healthcare systems. 

Summary 

The antibiotic pipeline, once one of the busiest areas of scientific research, has slowed severely in recent years due to a lack of resources and the issue of antibiotic resistance. But, with new models for antibiotic development utilising AI, there is large potential for a new ‘golden age’ of antibiotic development to dawn and antibiotic resistance to have smaller effects on both healthcare systems and individual patients’ lives. The idea of using AI to develop new antibiotics, despite the newfound excitement, is not a new one –Alto has covered this issue before including in this article. With new developments constantly being made, it’s to be expected that this topic comes up time and again. The hope is that each time it’s revisited it’s because we’ve gained ground in this critical area of research and worked towards finding solutions to one of the biggest issues in modern medicine. 

But should we make more antibiotics? The cost of drug development isn’t something to be overlooked, even when addressing an issue as important as antibiotic resistance – with experts such as Professor Chris Dowson at the University of Warwick asking the all-important question: “how can you make drugs that have no commercial value?”