**The Rise of the AI Scientists: How AI is Revolutionizing Scientific Research**
In 2010, Euan Ashley, a geneticist and cardiologist at Stanford University in California, led the first clinical analysis of a human genome, which took his team of 31 scientists nine months to complete. This week, while unpacking after a holiday, Ashley asked the AI tool Claude, developed by Anthropic in San Francisco, to examine his own genome to the same standard. The analysis took just 30 minutes and correctly identified an Alzheimer’s disease risk allele and gene variants affecting drug metabolism. Ashley had analyzed his genome in 2012 but had not published the results, noting that “there is no world in which this is not utterly remarkable.”
On June 30, Anthropic unveiled a platform called Claude Science, designed with biology research firmly in mind. This tool joins a department’s worth of general-purpose AI tools for science created by technology firms and academic laboratories. Others include offerings from OpenAI in San Francisco and Co-Scientist from Google DeepMind in Mountain View, California. Another is an open-source tool called Biomni, developed by academic researchers and described in *Science* just the day before. Researchers say there are many more emerging.
**What Are These AI “Scientists” and How Are Scientists Using Them?**
Sometimes called “AI scientists,” these tools are based on large language models that power chatbots, helping scientists with tasks such as literature reviews, data analysis, figure generation, and manuscript preparation. They are a form of agentic AI, in which requests are broken down into steps that often involve recruiting external software systems.
These scientific agents are distinct from more specialized research tools, such as the AlphaFold protein-structure-prediction model, but they can employ bespoke models. For example, Gabriele Corso, co-founder and chief executive of the London-based firm Boltz, and his team tasked a Claude agent to design an antibody that recognized two therapeutic targets, using the company’s open-source AI tools for protein-folding prediction and design. The AI’s outputs aligned with the protein designers’ intuitions; Boltz’s tools are among the dozens of specialized software systems that Claude Science and other AI scientists can interact with.
Clare Bryant, an immunologist at the University of Cambridge, UK, was an early adopter of Co-Scientist, which mines the scientific literature and other sources to come up with scientific hypotheses. Bryant, who was investigating immune responses to zoonotic pathogens, provided the tool with a grant application and further data. Some of the ideas it generated weren’t doable, but others were right in her lab’s wheelhouse. Her team is now testing an idea from Co-Scientist, introducing specific mutations into an innate-immune protein and seeing how they impact influenza infection. Bryant says she might have eventually come up with the experiment on her own, but it could have taken two years. “You feel like you’re talking to an oracle,” says Gary Peltz, a biomedical scientist at Stanford, who used Co-Scientist to identify existing drugs that could treat an organoid model of a disease called liver fibrosis.
**How Should Scientists Decide Which Tools to Use?**
Many scientists already use AI tools such as Claude to generate presentation slides and draft e-mails. But Ashu Singhal, president and co-founder of the cloud platform Benchling in San Francisco, estimates that less than 20% of labs have fully embedded AI scientists into their research. “It’s really important that people actually try these things out, rather than simply trusting what gets shared in headlines,” he says.
Singhal recommends that researchers trial several tools to work out which ones are suitable for which tasks. Hypothesis-generating AIs, such as Co-Scientist, might help during the earliest stages of a project. Later on, tools such as Claude Science and Biomni could carry out specific tasks, such as genomic data analysis. Corso recommends that researchers start with small tasks, the output of which can be verified easily. “Worst case, you have to do them over.”
**How Can Researchers Trust “AI Scientists”?**
As AI tools become more integral to research, trust and verification are paramount. Experts suggest that scientists:
– **Understand the limitations**: AI tools can generate hypotheses and speed up workflows, but they are not infallible and require human validation.
– **Verify outputs**: Especially in early stages, any AI-generated results should be tested and confirmed through experiments or data cross-checks.
– **Use multiple tools**: Comparing results across different AI platforms can help ensure reliability and uncover inconsistencies.
– **Stay updated**: The field is moving rapidly; keeping abreast of new tools and best practices is essential for responsible adoption.
—
### FAQ
**Q: What are “AI scientists”?**
A: “AI scientists” are advanced language models—often based on large language model (LLM) architectures—that assist researchers with scientific tasks like literature review, hypothesis generation, data analysis, figure creation, and writing. They operate as agentic systems that can plan and execute multi-step workflows, sometimes even using external tools or platforms.
**Q: How are scientists using AI tools today?**
A: Scientists are using AI tools for a range of tasks, from drafting emails and creating presentation slides to analyzing genomic data and designing experiments. For example, AI platforms like Claude, Co-Scientist, and Biomni help generate hypotheses, identify drug candidates, and interpret complex datasets. Adoption varies, with fewer labs fully integrating these tools, but early adopters report significant time savings and novel insights.
**Q: Are AI-designed proteins or molecules tested in real experiments?**
A: Some are. In the case of Boltz’s AI-designed antibodies, the outputs aligned with expert intuition, but experimental validation is still required. Several AI-generated molecules and therapeutic candidates have progressed to lab testing, but human oversight remains critical.
**Q: How can I choose which AI tool to use for my research?**
A: Start by identifying your task—whether it’s brainstorming ideas, analyzing data, or drafting text—and trial different tools to see which performs best. Begin with low-risk jobs whose results can be easily checked, and compare outputs across platforms to increase confidence in results.
**Q: How can I verify AI-generated results?**
A: Always validate AI outputs through experiments, literature checks, or independent computational methods. Treat AI suggestions as hypotheses rather than final answers, especially in early project phases.
—
### Conclusion
AI “scientists” are rapidly transforming the research landscape, offering unprecedented speed and efficiency in hypothesis generation, data analysis, and decision-making. Tools like Claude, Co-Scientist, and Biomni demonstrate the potential of agentic AI to augment human scientific effort—provided researchers approach them critically and thoughtfully. By understanding their capabilities, verifying their outputs, and integrating them strategically, scientists can harness AI as a powerful partner in discovery, accelerating innovation while maintaining rigorous scientific standards.
