AI System Runs DNA Experiments Without Step-by-Step Human Control
AutoDNA's demonstrated gains are research-stage, with no commercial product or timeline announced yet.
Updated on Nov. 4, 2026 in Artificial Intelligence
Live Poll
Do you think AI systems running scientific experiments will benefit society?
Shanghai Jiao Tong University researchers announced AutoDNA, an AI scientist system that designs and executes DNA experiments by following researchers' instructions, in a paper published in Nature. The system is research-stage: it was demonstrated in a laboratory experiment and has not been reported as a commercial product.
Why it matters
AutoDNA moves AI from analyzing experimental results toward running the experiments themselves, a shift that could compress the time between hypothesis and data in molecular biology. Its appearance alongside South Korea's government target for a fully autonomous AI scientist by 2035 suggests national programs are now treating autonomous research as a policy goal.
In a DNA storage experiment, AutoDNA ran 11,382 steps over 203 hours, encoding the sentence 'Intelligence is the ability to adapt to change' into 78 DNA strands, and it connected more than 20 types of equipment. Of 14 equipment errors during the run, the AI resolved 8 on its own while humans resolved 6.
The players
Shanghai Jiao Tong University
A leading Chinese research university whose team developed AutoDNA, an AI scientist system for automated DNA experimentation.
South Korea's Ministry of Science and ICT
The South Korean government ministry that set a 2035 target for an autonomous AI scientist under its K-Moonshot program.
The details
AutoDNA is built from several AI agents, each with a different role, based on Google's Gemini 2.5 Pro model. Rather than following a fixed script, the system repeatedly adjusted reaction times, reagent concentrations and washing methods to improve DNA synthesis, which is the chemical assembly of DNA one base at a time.
Timeline
May: South Korea appointed the inaugural head of its AI scientist mission.
End of September: South Korea's Ministry of Science and ICT appointed a new AI scientist mission head.
November 5: Shanghai Jiao Tong University announced AutoDNA in Nature.
2035: South Korea's target for an AI scientist to autonomously perform the full research process under K-Moonshot.
The Tech Race
AutoDNA is positioned as an AI scientist that controls physical lab equipment, a step beyond text-only reasoning models. South Korea's K-Moonshot program, with its 2035 goal for autonomous research, indicates that governments are now competing to field comparable systems.
AutoDNA remains a research demonstration and is not available to labs as a product, and no commercial timeline has been announced. Its near-term relevance is to researchers in DNA synthesis and molecular biology who may eventually work alongside systems like it.
The takeaway
The key trajectory is that AI systems are being built to run physical experiments, not only analyze data. Watch for whether AutoDNA's approach is reproduced on research tasks beyond DNA storage, and whether South Korea's 2035 autonomous-research target produces interim milestones.
Further reading
For broader context on how autonomous systems are entering research, explore Artificial Intelligence coverage.
Live Poll
Do you think AI systems running scientific experiments will benefit society?
