Autonomous Platforms Take Center Stage in Genetics
Autonomous and AI-driven laboratory systems have reached a new milestone as institutions in both China and the United States unveil sophisticated platforms designed to execute molecular biology workflows with minimal human oversight. On November 4, 2026, Nature reported that a team at Shanghai Jiao Tong University constructed an autonomous genetics laboratory called AutoDNA. The facility operates over 20 types of laboratory instruments using robotic arms and artificial intelligence agents to conduct complex experiments, including DNA synthesis and sequencing.
To say that someone has good hands is high praise in a molecular biology laboratory, where researchers must deftly handle delicate glassware, pipette samples, and manage intricate protocols. But in this unusual genetics lab in China, a team of robots driven by artificial intelligence is designing and executing experiments with no human hands involved in the physical handling. The facility represents one of the latest steps toward self-driving labs guided by artificial intelligence with minimal input from human scientists. With samples moved around by robotic arms, the Shanghai lab can operate more than 20 types of instruments, complete thousands of operations in a single project, and perform experiments that involve synthesizing and sequencing DNA among other tasks. Milad Abolhasani, a chemical engineer at North Carolina State University in Raleigh, notes that the system is remarkable both for its degree of autonomy and because it can readily incorporate a diverse range of lab equipment, stating that the future of science will involve a great number of autonomous labs.
This development follows a major announcement from AI firm Anthropic in September 2026. TechCrunch reported that Anthropic operates a wet biology lab in the Bay Area where its AI models analyze physical data. According to Anthropic, its AI assistant Claude directed 950 agents to search through data and successfully discovered a new enzyme system with properties reminiscent of CRISPR, performing tasks such as cutting, copying, and pasting DNA. The discovery was announced by the AI giant shortly after confirming the operation of its Bay Area wet biology facility, revealing a previously unknown enzyme system hidden in the DNA of bacteriophages, which are viruses that infect and replicate within bacteria.
Technical Architecture of the AutoDNA Facility
Traditional laboratory automation has long relied on machines executing isolated protocols, such as heating or mixing reagents. However, coordinating a complete workflow across multiple distinct instruments has typically required constant human intervention. According to Nature, the Shanghai Jiao Tong University team built AutoDNA from the ground up to integrate artificial intelligence into every phase of experimental design and execution rather than adding AI capabilities to a pre-existing system.
Haibo Chen, a computer scientist at Shanghai Jiao Tong University and co-author of the project, noted that molecular biology workflows involve intricate steps and high precision that traditionally demand expert configuration. In many existing systems, human experts are still required to configure the instruments, and artificial intelligence is mostly added to handle just one part of the process. AutoDNA overcomes these limitations by orchestrating disparate equipment to complete thousands of operations in a single project. During testing, the system successfully detected cancer-linked mutations and encoded textual quotations into DNA strands using nucleotide bases. Specifically, the researchers asked the system to encode a message in DNA using its four nucleotide bases, abbreviated as A, C, G, and T, selecting a popular quote often misattributed to cosmologist Stephen Hawking: Intelligence is the ability to adapt to change.
Industrial Implications and Biosafety Considerations
The convergence of artificial intelligence and physical laboratory automation introduces both significant scientific acceleration and complex oversight challenges. Milad Abolhasani told Nature that autonomous facilities represent the future of scientific experimentation across chemical and biological disciplines. Concurrently, the operational model adopted by companies like Anthropic highlights the ongoing debate over safety and human control.
While Anthropic’s AI model Claude identified a novel enzyme system in just 21 hours of concerted effort by searching through data using about 950 agents that burned through 210 million tokens, the physical experiments in the Bay Area facility were executed entirely by human scientists, as TechCrunch reported. Anthropic CEO Dario Amodei acknowledged that the discovery was based on the work of others and noted that a team from Stanford previously discovered a system that is in some ways similar to the one Claude found. Amodei and Anthropic emphasized that the discovery was found mostly, though not entirely, by Claude.
The revelation that Anthropic has been conducting biology experiments coincides with a period when AI CEOs, including Amodei, have publicly admitted that models have become so capable and potentially dangerous that the industry must slow down and develop safety-testing procedures. Amodei has stated that one of the things he fears most is that AI could be used for bioterrorism, while also expressing the belief that AI will cure most diseases in five to ten years. Consequently, Anthropic decided the rewards are worth the risks. The biology lab has not let Claude loose to control hardware directly; physical experiments were performed by human scientists. Amodei noted that the Bay Area lab looks like a typical molecular biology lab, conducting research that involves only the lower levels of the biosafety risk levels BSL-1 and BSL-2 without handling pathogens that can infect humans.
AI-powered biological research is also expanding across other academic and industrial institutions. Researchers from Stanford published a paper on their work with large language models and CRISPR, while researchers at the University of California, San Francisco, used AI to design enzymes from scratch. Google launched its first AI-powered biology research tool, AlphaFold, back in 2020. Despite current safety boundaries where human scientists execute physical protocols, Amodei has not ruled out a fully automated biology lab in the future, stating that eventually it may even be possible for Claude itself to safely perform the experiments by autonomously controlling lab equipment, with appropriate safeguards in place, even though such direct control is not practiced today.