In a development that has the research community paying close attention, Stanford scientists have used artificial intelligence to discover a naturally occurring molecule that activates GLP-1-like pathways in the body - without the gastrointestinal side effects that plague current drugs.
The study, published in late July and covered by ScienceDaily, represents a fundamentally different approach to the GLP-1 problem. Instead of designing synthetic compounds that activate the GLP-1 receptor, the Stanford team asked a different question: does the human body already produce something that does this job naturally?
The Discovery
The Stanford team used AI to screen the human proteome - the entire set of proteins expressed in the human body - looking for molecules that could interact with the GLP-1 receptor or related metabolic pathways. The AI identified a naturally occurring molecule that had not previously been linked to GLP-1 signalling.
What makes this discovery notable is not just the molecule itself, but the profile it showed in early testing. The naturally occurring compound appeared to activate metabolic pathways associated with glucose regulation and appetite suppression while largely bypassing the mechanisms that cause the nausea, vomiting, and gastrointestinal discomfort that are the most common side effects of GLP-1 drugs.
The researchers described the compound as a “natural Ozempic” - a comparison that immediately caught media attention.
Why This Matters for the GLP-1 Research Space
The side effect profile of GLP-1 drugs is one of the most significant barriers to wider adoption. Studies consistently show that 20-40% of people who start GLP-1 therapy experience significant nausea, particularly during dose titration. Some people discontinue treatment entirely because they cannot tolerate the gastrointestinal effects.
A molecule that preserves the metabolic benefits without the side effects would represent a genuine breakthrough - not just for patient comfort, but for adherence rates, treatment duration, and ultimately, outcomes.
The Stanford finding also raises a broader question: how many other naturally occurring regulatory molecules are we missing? The GLP-1 system was discovered through traditional biochemical approaches in the 1980s. The idea that there might be parallel or complementary systems that we simply did not have the tools to find is both humbling and exciting.
The AI Angle
This study is part of a wave of AI-driven drug discovery that is reshaping how researchers find candidate molecules. The traditional approach - screening thousands of synthetic compounds, testing them in cell assays and animal models, and gradually refining leads - is slow and expensive.
AI approaches can search vastly larger chemical and biological spaces much faster. In this case, the AI was searching the naturally occurring human proteome, not a library of synthetic compounds. That is a meaningful distinction - the molecule the AI found is something the human body already produces, which means it has already passed some basic biological compatibility tests.
The Stanford team used a machine learning model trained on protein-protein interaction data and receptor activation patterns. The model identified candidates that the researchers would not have thought to test, including the molecule that ultimately showed the promising side effect profile.
What the Research Says
The findings are at an early stage. The AI identified the molecule, and initial cell-based and animal studies showed promising results. But human trials have not been conducted. It is a long way from a computational screen and early preclinical data to a viable therapeutic.
However, the approach itself is noteworthy. AI-guided discovery of naturally occurring metabolic regulators opens up possibilities that extend well beyond GLP-1. If the same approach can be applied to other receptor systems, we could see a wave of natural ligand discoveries across multiple therapeutic areas.
For the Australian research community, this is a reminder that the peptide landscape extends beyond the well-known GLP-1 analogues. Nature has been doing molecular discovery for billions of years. We are only now developing the tools to understand what it has already produced.
Sources
Source: ScienceDaily - AI helps Stanford scientists discover ’natural Ozempic’ without the usual side effects - published July 24, 2026
Source: Stanford Medicine - GLP-1s 101: What the science says about weight loss, side effects, safety - published June 29, 2026
Source: ScienceDaily - Oral GLP-1 drugs may quiet the brain’s food craving circuit - published July 25, 2026
Source: Nature - Genetic predictors of GLP1 receptor agonist weight loss and side effects - published April 8, 2026
Related Reading
For more on the science of GLP-1 receptor agonism, see our GLP-1 weight loss science explainer. For the latest on how current GLP-1 drugs compare, check our obesity drug landscape coverage.
Join the Grey Highway Telegram community to discuss this research with other Australian researchers.
This article is for educational and informational purposes only. It does not constitute medical advice, therapeutic recommendations, or endorsements of any compound. Grey Highway is a research-education community. We do not sell, supply, or promote the use of research compounds. Always consult a qualified healthcare professional regarding health decisions. For Australian regulatory information, visit the TGA website.