Seventeen countries have just adopted a declaration in Kyoto aimed at making massive use of artificial intelligence to accelerate scientific discoveries.... The Kyoto Vision aims to develop more autonomous laboratories, facilitate access to data and computing power, and revise research funding. The United States, Japan, the United Kingdom, Germany, and South Korea are among the signatories. France, however, does not appear on the official list. What does this document actually entail, and can AI truly automate a portion of scientific research?
Key points to remember
- The United States and sixteen other countries have adopted the Kyoto Vision for a Golden Age of Science October 4, 2026.
- The text encourages the use of AI in scientific work and automated discovery systems.
- He calls for broader access to models, data, computing infrastructure and experimental equipment.
- France is not among the 17 signatory countries; the document does not explain the reason.
- The declaration does not specify a budget, deadline, or precise legal obligation.
- The expression "super intelligence" used in the text does not demonstrate the current existence of an AI superior to humans.
What is the Kyoto Vision?
The Kyoto Vision for a Golden Age of Science is an international declaration presented on October 4, 2026 during the Science and Technology in Society Forum held in Kyoto, Japan.
It is supported by Argentina, Bulgaria, Chile, Cyprus, Germany, Greece, Indonesia, Italy, Japan, Kazakhstan, South Korea, New Zealand, Poland, Singapore, the United Arab Emirates, the United Kingdom and the United States.
The statement starts from an observation: artificial intelligence can already help researchers to sift through large quantities of publications, build more detailed models, and automate certain steps of experimentation.
The signatories want to encourage this development while reforming the way research is organized, funded and evaluated.
How can AI accelerate a scientific discovery?
Today, a researcher often has to consult hundreds of articles, select hypotheses, design an experiment, analyze the results, and then decide on the next step. Some of these tasks can already be assisted by AI models.
A system can, for example, accomplish several missions.
- quickly analyze a vast amount of scientific literature; ;
- identify relationships between results from different disciplines; ;
- to propose molecules or materials to be tested; ;
- to simulate a physical or biological phenomenon; ;
- to operate certain laboratory equipment; ;
- analyze the results and propose the next experiment.
When the system chooses a hypothesis, launches an experiment, measures the result, and then automatically adjusts the next step, we speak of closed-loop discovery. The goal is no longer just to help the scientist write or calculate, but to automate part of the experimental cycle.
What would an autonomous laboratory look like?
An autonomous laboratory combines several technologies, including artificial intelligence, robotics, connected scientific instruments, and analytical software.
In medical research, AI could select promising molecules while robots perform tests. The results would be automatically analyzed to determine which variants to study next.
In materials science, a system could compare thousands of compositions to search for a more durable battery, a lighter material, or a new superconductor.
In the fields of energy and climate, AI can improve simulations, optimize chemical reactions, and help design components. In biology, it can combine genetic, cellular, and clinical data that are difficult for a single team to process.
These systems do not necessarily eliminate the researcher. They can primarily increase the number of hypotheses tested and reduce the time required between experiments.
Kyoto Vision also wants to change the way science is funded
The text does not deal solely with technology. It also proposes experimenting with new ways of funding research, including long-term grants, rapid funding, prizes, and scientific challenges.
He highlights "metascience," a discipline that studies how research is organized and funded in order to identify what truly fosters discoveries.
The signatories also state that they want to reduce certain administrative burdens and provide earlier support to students or young researchers deemed promising.
But these intentions are not accompanied by any common amount, any precise timetable, and any detailed monitoring mechanism.
Why is France not among the signatories?
France is not included in the official list published by the White House. Spain, Canada, and several other major scientific powers are also absent.
None of the documents consulted explain this absence. Therefore, it does not allow us to conclude that France refuses the use of AI in research or that it definitively rejects the declaration.
France already has laboratories, universities, public programs, and companies working on scientific AI. The question is rather whether it will later join this initiative, propose a different framework at the European level, or prioritize its own programs.
«Super intelligence»: an expression to be used with caution
Kyoto Vision uses the term "super intelligence" instead of artificial intelligence. This vocabulary aligns with the terminology currently advocated by the Trump administration.
This does not mean that current scientific systems understand all domains better than humans. An AI can excel at data analysis or predicting patterns, but it can still produce an erroneous conclusion if the data is incomplete or the experiment is poorly designed.
The official statement also emphasizes reproducibility, transparency, communication of uncertainties, peer review and the value of negative results.
What are the main risks?
Accelerating research does not automatically guarantee better discoveries. Several risks must be taken into account.
- biased or poor quality data; ;
- hypotheses generated without sufficient scientific basis; ;
- a concentration of tools among a few large companies; ;
- high computing and equipment costs; ;
- results that are difficult to reproduce; ;
- dangerous uses in biology, chemistry or cybersecurity; ;
- an excessive reliance on template and cloud providers.
An autonomous laboratory must therefore maintain validation mechanisms, access limits and clearly identified human responsibility.
What can this statement change?
In the short term, the Kyoto Vision does not create a single global program. It primarily constitutes a political signal addressed to researchers, investors, and administrations.
It can foster new partnerships, accelerate funding for AI-powered scientific platforms, and encourage infrastructure sharing. It can also intensify competition among countries to attract researchers, computing power, and deep tech companies.
Its effectiveness will depend on the decisions taken after signing, including the budgets actually committed, access to data, construction of laboratories, international cooperation and transparent publication of results.
Our reading
Kyoto Vision highlights a crucial development: the next major battle in artificial intelligence will not be fought solely in chatbots. It will also take place in laboratories, hospitals, research centers, and industrial infrastructures.
AI can help test more hypotheses and accelerate certain discoveries. But a one-page document isn't enough to create a scientific golden age. The lack of a budget, deadlines, and binding mechanisms means that the Kyoto Vision should be viewed as a political direction, not as a revolution that has already taken place.



