Reforms needed at The Alan Turing Institute, finds review

The quinquennial review was carried out by a panel of independent experts and sought to outline the value of The Alan Turing Institute’s activities and outputs during the first five years of its operation and assess its future strategy.

The findings of the panel’s review come after The Alan Turing Institute was provided with £100m investment in the UK government’s spring budget to help address national and international challenges in AI and data science.

The review said there had been “clear value” in the work carried out by the national institute for data science and AI and said there was a “positive outlook” for its future and there was “confidence” in the approach adopted by the incoming chief executive officer and board of trustees.

However, the review also highlighted five areas of concern, broadly covering governance, implementation of strategy, relationships with the ecosystem, financial management and operational effectiveness.

For governance, the review panel said that while there is evidence that The Alan Turing Institute played a national role in the AI ecosystem, some UK research stakeholders felt disenfranchised and “the potential of the institute to act on behalf of the whole ecosystem has not yet been achieved, a key part of the national institute role”.

It recommended the institute aim to “represent all actors in the UK ecosystem on an international stage, and enable research that is additive to, not duplicative of, that carried out in other institutions”.

In addition, the review said there was a clear need to evolve the governance and leadership structure of the institute to better reflect the diversity in the wider industry.

On implementation of strategy, the panel backed The Alan Turing Institute’s refreshed strategy and vision, as well as focus on defence and national security, environment and sustainability, and transformation of healthcare.

The report recommended the institute create and action a clear business plan and delivery mechanisms to achieve its stated vision.

When discussing relationships with the ecosystem, the report said that the institute “acts as a marquee investment for the UK and has an internationally recognised brand”, as well as building an extensive network of partners and delivering leading research.

However, the report added that “some stakeholders interviewed identified challenges engaging with the institute, such as insufficient opportunities to work with the institute, perceived operational barriers to entry, and seeing the institute as a competitor rather than a collaborator”.

The panel recommended a clear statement of how actors within the AI ecosystem interact, to avoid duplication, ensure complementarity of function, realise the benefits of coordinated action and provide simplicity of navigation.

Financial oversight was “not sufficient” to provide assurance as to how all public funding is spent within the institute, the report added.

There is disagreement as to what the institute’s role should be and how it should interact with academia, the report said, as well as a need to better define how the institute should feed into and undertake work on behalf of government and the public sector, as well as working with businesses.

The report concluded that any future core funding from UK Research and Innovation (UKRI) and the EPSRC should therefore be conditional on resolution of the specific concerns, to ensure the institute acts as a national institute for AI and data science.

This centred on four recommendations covering a revised governance structure for the institute, clarity on the primary activities of and anticipated outcomes of the ‘Turing 2.0’ strategy, a business implementation plan and a clear financial plan covering the next five years.

Dr Kedar Pandya, executive director for cross-council programmes at EPSRC, said: “The review recognises the excellent work done by the Turing since its inception and identifies the clear value that a national institute provides within the UK’s AI research and innovation ecosystem.

“The review will help to shape the institute’s next phase, which will see it play an important role in contributing to the government’s priorities and utilising AI and data science to drive forward research, economic and societal impacts.

“We are working with the Turing to ensure recommendations made by the review are implemented, to maximise the benefits of this investment.”

Jean Innes, chief executive at The Alan Turing Institute, said: “I’d like to thank the QQR panel for their thorough review and thoughtful recommendations and advice for the next phase of the institute, including their recommendation for a further five years of funding.

“This provides a foundation for longer term planning as we chart a path to delivering maximum impact.

“We greatly look forward to working with partners to harness the collective talent of the community and push the boundaries of data science and AI for the public good.”