Strategy focuses on five areas of ‘Britain’s existing strengths’, with £137m of funding available
The government has set out plans to accelerate the development of AI-driven science in the UK, including funding for “autonomous” labs, making AI-enabled datasets compulsory and expanding PhD training for “AI-savvy” researchers.
In its strategy for the use of AI in science, published on November 20, the government also said it would provide more funding for research into the implications of the growing role of AI in research and ensure that its introduction does not have a “negative impact on the integrity of research”.
Big technology companies such as OpenAI and London-based Google DeepMind are racing to develop AI tools specifically designed to speed up research, and the United States, China and the European Union have announced strategies on how to leverage AI for scientific leadership.
In a joint foreword to the UK strategy, Science Minister Patrick Vallance and Artificial Intelligence Minister Kanishka Narayan said: “If we act quickly, we have a chance to dramatically increase scientific productivity and establish UK leadership at a time of unprecedented scientific innovation. If we move too slowly and without a unified approach, our scientific institutions could be left behind, overtaken by more ambitious and nimble emerging leaders.”
Accelerating breakthrough
The strategy includes £137m of funding from the government’s £2bn wider AI agenda budget for 2026-30 to “accelerate AI-led scientific breakthroughs in the UK’s priority areas” and includes “long-term funding”. [for] Leading researchers focused on new organizational structures. ”
The strategy specifies five broad priority areas “identified based on the UK’s existing strengths” in line with the government’s industrial strategy: engineering biology, fusion energy, materials science, medical research and quantum technology.
In addition to the priority areas, the strategy also launches “missions” for AI for science, the first of which is accelerating drug discovery to “develop clinically viable medicines within 100 days by 2030.” A further mission will be selected by the Department for Science, Innovation and Technology (Dsit) and UK Research and Innovation (UKRI) in 2026.
autonomous lab
As part of efforts to scale up AI-driven science in the UK, the government’s Sovereign AI Department will launch a funding appeal “seeking proposals to develop or expand autonomous laboratory platforms in the UK”. Such an autonomous platform can analyze the results and be used to control further experiments.
The government said it would also support teams building more ambitious projects, such as those being considered under the Agency for Advanced Research and Inventions’ AI Scientist Funding Programme, which looks at the potential of systems that “can carry out the entire research process without ongoing human intervention”.
research integrity
Meanwhile, the government said it was important that “the introduction of AI does not adversely affect the integrity of research” and pledged to fund research into the implications of incorporating AI into scientific discovery.
The company said it was “acutely aware of the importance of responsible AI implementation in science and beyond,” and said the prospect of AI automating scientific discovery “is disruptive and poses a significant challenge to many current research practices.”
The UK Metascience Unit, jointly run by Dsit and UKRI, is to launch a national survey to understand the uptake of AI across UK research.
AI-enabled data
Making research data AI-enabled is also a key element of the Government’s strategy, and UKRI will take steps to ensure the UK’s national laboratories and scientific facilities lead the way in storing and managing data to enable AI-enabled research.
Funding agencies are also expected to update their policies to ensure that the research they fund “supports modern AI-enabled data environments.”
The strategy also outlines efforts to collect so-called ‘dark data’ – negative results from failed or inconclusive experiments, or data that does not support publications – and UKRI will launch a number of pilot approaches.
specialized training
The government also plans to invest in training research talent, especially those in the early stages of their careers, to use AI in science. It said it would expand its offer of AI for science PhD training “with the aim of starting the training of at least 1,000 researchers who are familiar with AI and domain science or apply AI in their research.”
Additionally, the strategy promises to invest in “the full range of research and technology experts” and says the government will convene universities and research institutes to develop technical career paths.
