The University of Manchester has announced its selection as the second UK ELIAS node, with three research themes and plans for European collaboration and doctoral training.
A new European research connection
The University of Manchester has been selected to join the European ELIAS network as its second UK node, alongside Cambridge. Announced by the University on 21 October 2026, the development connects Manchester’s AI research with a European structure for academic, industry and doctoral collaboration.
Professor Sophia Ananiadou leads the Manchester ELIAS node, which will be hosted at the Christabel Pankhurst Institute. Its announced programme centres on three research themes, supported by existing university expertise and plans for exchanges with other European research centres.
The cover and article images are AI-generated illustrations of generic scenes related to this topic.
Manchester becomes the second UK ELIAS node
The University’s announcement followed the selection announcement in Stuttgart on 8 October 2026. The Alliance published its expansion release on 9 October. These are separate publication and announcement dates, rather than three different stages of operational delivery.
Manchester became the second UK ELIAS node through an expansion that added eight regional nodes. According to the dated Alliance release, this doubled the network to 16 nodes across nine countries.
That total updates the earlier network of eight nodes. The original group launched in November 2024 in Amsterdam, Barcelona, Cambridge, Copenhagen, Munich, Potsdam, Tübingen and Zurich.
Selection followed a competitive call launched in May 2026. The assessment considered scientific quality, institutional backing and operational capacity. It also examined engagement with non-academic advisers and deep-tech entrepreneurs: people developing businesses around substantial scientific or technological advances.
How the European ELIAS network connects research with innovation
ELIAS stands for European Leadership in Innovation with AI and Science. The ELIAS Alliance works closely with ELLIS, the European Laboratory for Learning and Intelligent Systems. The similar names describe related organisations, not interchangeable labels.
The Alliance’s nodes are intended to connect academic research with industry. Their remit includes supporting researchers who want to explore entrepreneurial ventures and strengthening education around open-source foundation models and creating value from research.
Foundation models are models trained broadly enough to support different tasks or further specialist development. The emphasis on open approaches concerns how researchers can share and build on that work, rather than each team starting independently.
The wider ELIAS research mission covers computational efficiency, modelling the societal effects of policy and understanding impacts on individuals. Sustainability, social cohesion and individual rights form part of that stated mission alongside research and industry applications.
Several organisational layers sit behind this work: the Alliance’s regional nodes, the wider research consortium and its Virtual Centre of Excellence. Their membership totals describe different groupings and should not be combined.
The University of Manchester was already listed as an associated partner in the wider consortium. Establishing the Manchester ELIAS node is therefore a distinct development, not evidence that the University had no previous connection with ELIAS. The new announcement gives that relationship a regional innovation structure.
The three research themes planned for Manchester
Manchester’s contribution to the European ELIAS network is organised around three announced research themes. They connect methods for developing AI with questions about evidence, reliability and scientific understanding.
Trustworthy AI, digital twins and computational medicine
The first theme combines trustworthy AI, digital twins, in-silico trials and regulatory-grade evidence. It draws on multimodal learning, causal AI, simulation and computational medicine.
A digital twin is a computational representation of a real-world system. In-silico trials use computer-based methods, while multimodal learning works across different forms of data. Causal AI examines cause-and-effect relationships rather than relying only on patterns of association.
Open foundation models, agents and reliability
The second theme covers open foundation models, agentic and trustworthy AI. Its scope includes language technologies, specialist models, multilingual systems, intelligent agents, and reliability and safety evaluation.
Agentic AI concerns systems that can organise and carry out sequences of tasks. Here, that sits alongside research into how systems perform and how their behaviour can be evaluated. The theme therefore covers both capability and the evidence needed to assess it.
Scientific discovery and complex systems
The third theme concerns scientific discovery and complex systems. It brings together probabilistic machine learning, optimisation, uncertainty quantification, scientific machine learning and human-centred AI.
Probabilistic methods represent uncertainty, while uncertainty quantification examines how confident researchers can be in a result. Optimisation seeks solutions within defined objectives and constraints. Together, these approaches address how AI can support scientific reasoning without treating every output as certain.
The existing research behind the new node
The second UK ELIAS node builds on several existing Manchester research structures. The University names the National Centre for Text Mining, known as NaCTeM, ELLIS Manchester, the Centre for AI Fundamentals, existing doctoral training and the Christabel Pankhurst Institute.
NaCTeM provides a concrete example of the language-technology expertise behind the announcement. It supplies text-mining services for the UK academic community using natural language processing: computational methods for working with human language.
The centre’s text-mining tools and research resources include software, publications, demonstrations, tutorials and research events. Three examples show the different kinds of task involved in extracting useful information from large collections of written material.
TerMine identifies key phrases in text. RobotAnalyst supports the identification of studies for systematic reviews, helping reduce the human workload in that process. AcroMine finds expanded forms of acronyms in MEDLINE, a biomedical literature database.
These examples address different parts of handling research literature: identifying terminology, finding relevant studies and interpreting abbreviated language. They give substance to the University’s reference to language technologies without requiring assumptions about future node projects.
They also establish an important distinction in the chronology. This is existing text-mining work, not a collection of products newly launched by the Manchester ELIAS node.
Doctoral collaboration and European research exchanges
The Manchester ELIAS node plans research, exchanges and doctoral collaboration with Cambridge, Finland, Franconia and Barcelona. It also intends to contribute to the Alliance’s developing Innovation PhD initiative.
That initiative must be distinguished from the established ELLIS doctoral programme. The existing programme provides useful context for European research training, but it is not evidence of a newly opened Manchester recruitment scheme.
The ELLIS PhD and postdoctoral programme information describes a central application process connecting researchers with European machine-learning and AI laboratories. Training and networking include summer schools, workshops and bootcamps, alongside links with industry.
Joint supervision is a substantive part of the programme. Its arrangements include at least six months of research at another laboratory or with an industry partner. A mobility fund supports long- and short-term visits.
The programme also describes an interdisciplinary track linking AI with fields including law, biology, social sciences and humanities. Supervision pairs an ELLIS Fellow or Scholar with a tenured researcher whose expertise lies outside mainstream machine learning.
The supervisors may work in the same location or in different countries. Interdisciplinary supervision therefore concerns the combination of expertise, not simply international travel.
Manchester’s proposed contribution to the Innovation PhD approach sits alongside this established framework. The developing initiative explores links between doctoral research, practical challenges and entrepreneurship.
The node announcement does not specify available Manchester doctoral places, application deadlines, eligibility or stipends. For prospective researchers, its significance is the intended direction of collaboration. Any particular recruitment opportunity would need its own published terms rather than being inferred from selection as the second UK ELIAS node.
Regional partners and the route towards innovation
The University identifies six supporters: Unit M, University of Manchester Innovation Factory, Turing Innovation Catalyst, Industrial Biotechnology Innovation Catalyst, Bionow and the Cambridge–Manchester Innovation Partnership.
The wider European ELIAS network provides context for the intended connections between research and innovation. In its 9 October expansion announcement, the Alliance reported that the ETH AI Center-led AI Launchpad had connected ten European innovation hubs and supported more than 20 startups by August 2026.
Launchpad activities help research-led teams connect with mentors, industry contacts, markets and investors beyond their home region. Those figures describe European activity before Manchester’s selection announcement; they are not measured outcomes for Manchester.
The Alliance is also developing an Open Pipeline for more collaborative, reusable foundation-model research. Alongside the Innovation PhD approach, this shows two different routes towards its aims: sharing research foundations and connecting researcher development with practical challenges.
For Greater Manchester, the immediate significance is a new institutional connection linking existing expertise with European collaborators and named regional supporters. It is not a quantified forecast of employment, investment or business growth.