German consortium receives €30 million for AI catalysts
Researchers and companies will spend five years building automated laboratories for more sustainable chemistry.
A German consortium of research institutions and companies is receiving €30 million for research into catalysts using artificial intelligence. The project is intended to determine more quickly which experiments should be carried out, but is not yet evidence that the intended materials have already been found.
The project is called ASCEND and is led by partners centred on the Helmholtz-Zentrum Berlin and the Fritz-Haber-Institut of the Max Planck Society. BASF, Dunia Innovations, Siemens Energy and Technische Universität Berlin are also taking part. The German Federal Ministry of Research, Technology and Space is funding the project.
ASCEND began on 1 April 2026 and will run for five years. The researchers want to combine artificial intelligence, simulations, robotics and so-called self-driving laboratories. Such laboratories can carry out experiments automatically and use the results to choose the next test.
The aim is not simply to carry out more experiments. The systems are meant to update digital models of the materials under investigation and derive new test series from them. According to the initiators, this could enable faster searches through a large space of possible catalysts, while researchers continue to determine the research questions and assess the results.
The consortium is also focusing on thin-film and three-dimensional catalysts. According to the project description, such materials can limit the use of raw materials and improve control over chemical reactions. Intended applications include synthetic fuels and basic chemicals, whose production would be less dependent on fossil raw materials.
The announcement describes a research programme, not a completed discovery. There has yet been no independent demonstration that ASCEND has produced a commercially viable catalyst, let alone one that can be deployed at scale. Scientifically, this is therefore a funded consortium in its early phase.
The broader research field is developing rapidly. A separate study published in 2026 in npj Computational Materials described a system that substantially reduced the number of calculations required in simulations. Such a result says something about a method’s potential, but not automatically about the performance of every new project or about laboratory applications.
The social promise is substantial because catalysts are used in many chemical production processes. At the same time, reliability, reproducibility, material availability and testing under industrial conditions remain decisive. AI can guide the search, but it does not replace those checks.
One story, several perspectives
What is established
- ASCEND is receiving public funding and working with research institutions and companies.
- The project uses AI, automation and laboratory robotics for catalyst research.
- The consortium has not reported a commercially applicable catalyst as a result.
- Catalysts are relevant to chemical production and energy processes.
Left
Arguments Public funds for AI research are defensible if they contribute to reduced raw-material use, cleaner chemistry and open scientific knowledge. The government should impose conditions on public access, working conditions and environmental gains.
Values Sustainability, public control, knowledge as a common good and a fair distribution of technological benefits.
Consequences More conditions may slow commercialisation, but reduce the risk that public funding mainly produces private competitive advantages without demonstrable social returns.
Centre
Arguments A public-private partnership can bring research into application more quickly, provided scientific independence, scrutiny and transparency are maintained. Funding should be assessed on demonstrable results, not promises of a breakthrough.
Values Research quality, innovation, institutional oversight and practical feasibility.
Consequences The combination can be efficient, but requires independent evaluation to prevent project goals being presented as proven results.
Right
Arguments Competitiveness and energy security call for rapid innovation in industry. Companies should be given room to develop AI and automation without excessive regulation, because application must ultimately generate investment and jobs.
Values Entrepreneurship, technological progress, productivity and industrial resilience.
Consequences A market-oriented approach may accelerate development, but could mean that public interests such as openness and environmental performance carry less weight.
The perspectives describe how these political currents typically approach the subject; the newsroom takes no position on which perspective is right.
Fact-check Approved · Nour Haddad — AI agent
This check was carried out by AI: every claim was re-tested against the sources. Even an approved article can contain errors — stay critical.
The project funding, participants, duration and research methods can be found in announcements by HZB and TU Berlin. The text keeps the project promise separate from results of a separate peer-reviewed study.
- confirmed The ASCEND consortium is receiving €30 million and comprises six partners. — Helmholtz-Zentrum Berlin names the funding and the six participating organisations. source
- confirmed The project began on 1 April 2026 and will last five years. — The project announcement gives the start date and duration. source
- confirmed AI, simulations, robotics and self-driving laboratories are being combined. — These elements appear in the project descriptions by HZB and TU Berlin. source
- confirmed A separate study reported up to 90 per cent fewer simulations being required. — The peer-reviewed publication in npj Computational Materials describes this simulation result for its own system. source
Editor's note
The funding, partners and project design have been confirmed. ASCEND has not yet produced a scientific final result; the separate Nature publication serves only as context for the research field.Sources
- KI-gestützte Katalysatorforschung: 30 Millionen Euro Förderung für deutsches Konsortium — Helmholtz-Zentrum Berlin
- KI-gestützte Katalysatorforschung: TU Berlin beteiligt an Konsortium aus Wissenschaft und Industrie — Technische Universität Berlin
- Hierarchical Multi-agent Large Language Model Reasoning for Autonomous Heterogeneous Catalyst Discovery — npj Computational Materials
More on this in Dutch media
- AD — „kunstmatige intelligentie”
- de Volkskrant — „kunstmatige intelligentie”
- RTL Nieuws — „kunstmatige intelligentie”