Scientists in Sweden have taken a significant step towards automating the research process with the development of an AI system that can generate hypotheses, design experiments, and interpret results with minimal human intervention. The work, carried out by researchers at Chalmers University of Technology in Gothenburg and the University of Gothenburg, combines large language models, extensive biological databases, and robotic laboratory equipment into a closed-loop system that mimics the entire scientific method.
The system was tested on the yeast Saccharomyces cerevisiae, commonly known as brewer's or baker's yeast. It analysed a database of roughly 60,000 phenotypical, physiological, and metabolic relationships, and from that data generated nearly 2,000 testable predictions about how nutrients affect cell growth and stress resistance. The AI then selected the most promising hypotheses, designed comparison controls, and converted the ideas into step-by-step instructions that laboratory robots could execute.
Once the experiments were completed, the AI analysed the results, determined whether its predictions were correct, and automatically updated or refined any hypotheses that had been refuted. This iterative process allows the system to learn from each cycle and improve its understanding of the underlying biology.
“It is too much information for a human to analyse, but our AI scientist could identify promising biological questions, recommend experiments to test them, evaluate experimental outcomes and iteratively refine its understanding based on new evidence,” said Ievgeniia Tiukova, a researcher at Chalmers and one of the authors of the study.
A new kind of research partner
The researchers emphasise that this is not just a decision-support tool. “Rather than serving solely as decision-supporting tools, the AI scientist actively generates new scientific knowledge,” Tiukova said. The system is designed to work alongside human scientists, accelerating the pace of discovery in fields such as biology, medicine, and biotechnology.
Ross King, senior author of the study at the University of Gothenburg, believes that AI scientists will become valuable collaborators. “AI Scientists will collaborate with human scientists to accelerate discoveries across biology, medicine and biotechnology,” he said. “Such AI systems have the potential to reduce the time required to explore complex scientific questions, and optimise the use of laboratory resources.”
The potential benefits extend beyond speed. Automated systems could minimise human bias, reduce variation caused by human error, and ensure that experimental protocols are recorded completely and consistently. This could improve the reproducibility of scientific results, a long-standing concern in many research fields.
Humans remain essential
Despite the advances, the authors are clear that human oversight will remain crucial. “Humans will remain essential for setting priorities, interpreting broader contexts, developing research programmes and ensuring compliance with ethical rules,” King said. He added that future generations of autonomous laboratories will become increasingly capable of collaborating with humans and will be valuable partners in tackling the most challenging questions in biology and medicine.
The development comes as international bodies begin to address the governance of AI in research. The World Health Organization recently published recommendations for researchers, ethics committees, regulators, funders, and policymakers to ensure that AI-enabled health research is conducted responsibly and for the benefit of all. “Artificial intelligence is creating unprecedented opportunities to accelerate health research and improve people's lives,” said Meg Doherty, director at WHO’s Department of Science for Health.
While the Swedish team's work is still in its early stages, it represents a notable milestone in the move towards automated scientific discovery. As robotics and AI continue to improve, the vision of a fully autonomous laboratory—one that can generate and test hypotheses around the clock—is moving closer to reality. For now, the human scientist remains at the centre, but their role is evolving from hands-on experimenter to strategic partner in a new era of AI-assisted research.
