the government’s newly published Green (consultation) Paper, Invest 2035: The UK’s Modern Industrial Strategy (https:// tinyurl.com/odiNDL4), has a goal of “using public sector data as a driver of growth”, we must ensure that the approach to implementing the library takes account of the varying data maturity of public sector and research organisations whose data might come within its remit. In the ODI’s recommendations for the AI Action Plan (https://tinyurl.com/odiNDL1), we suggest that the library initially focuses on three key areas: high-quality public data, federated Trusted Research Environments (TREs), and cultural heritage data. These datasets offer significant benefits and are well-structured for reliable public-sector deployment. Making them available to digital innovators could enhance public service delivery and stimulate the ecosystem necessary for sustainable impact. This would also give the NDL robust foundations on which to build. Regardless of the roadmap to building the library, resources and investment will be needed, just as they are required for building physical infrastructure. In some ways, as our Senior Policy Advisor, Gavin Freegard, noted back in July the NDL can be likened to a traditional library. He cited a report by Onward proposing that “the Government should establish a British Library for data – a centralised, secure platform to collate high-quality data for scientists and start-ups.” Just as libraries store, organise, and provide access to knowledge, the NDL can curate and make data accessible to researchers, businesses, and public services. However, just as libraries have different access levels depending on the sensitivity of their materials, the NDL will need to offer tiered access to protect personal data. Data stewards (https:// en.wikipedia.org/wiki/Data_steward) will be


28 INFORMATION PROFESSIONAL


critical in guiding users through this vast resource, ensuring the NDL becomes a trusted and effective driver of innovation and societal benefit. Indeed, building and maintaining public trust will be a critical success factor in the project. The Lloyd’s Register Foundation’s World Risk Poll 2024 reveals that only 11 per cent of global citizens trust their governments to protect personal data, and the problem is particularly acute when it comes to personal or sensitive data, like benefits information or health records. The 2021 GPDPR campaign (https://tinyurl. com/odiNHSdata), where patient opt-outs spiked, shows the risks of ignoring public concerns. In building the National Data Library, we can avoid this by exploring the potential of new Privacy Enhancing Technologies (PETs) (https://tinyurl.com/ odiPETS1), which include personal data stores, federated learning models and multi- party computation. These technologies, which the ODI is researching, offer ways to ensure data security and empower individuals to control data about them. For example, in a Federated learning


(https://en.wikipedia.org/wiki/Federated_learning) model, the model travels to the data rather than the data being ‘pooled’ by the user for the algorithm to be deployed. The model is trained locally on each device; the data never leaves its original location. This technology is already being successfully used in projects such as the EU’s MELLODDY (www.melloddy.eu) and the secure healthcare research collaboration between Moorfields Eye Hospital (https:// tinyurl.com/odiMoor) and Bitfount (www. bitfount.com/), which aims to improve the early detection of eye diseases. Enabling algorithms to be trained across multiple local datasets without exchanging the underlying data presents great potential to unlock value from data that might traditionally have been kept closed. With personal data stores, individuals control their personal data and decide how


it can be used across the Web. There are several examples of this, including Solid (https://solidproject.org). Solid is an open- source project, community and standard (or protocol), originated by the ODI’s co- founder Sir Tim Berners-Lee. Stewarded by the ODI since October 2024, Solid’s model is already proving effective. For example, in Flanders, a company called Inrupt (www.inrupt.com), has provided enterprise-grade Solid software for projects, including with the government of Flanders (https://tinyurl.com/odiFland) – through the Flanders’ Data Utility Company (https://vimeo.com/709214424). This has given 6.8 million citizens their own Personal Online Data Stores (Pods) through which they can share data with government services, demonstrating the potential for trusted, transparent public sector data exchanges.


Multi-party computation (MPC) is a cryptographic protocol that enables individual stakeholders who hold sensitive data to pool it with others for joint computations without revealing the underlying data itself. Again, this offers a potential solution to some of the challenges around sharing sensitive public sector information via the NDL. Earlier this year, the ODI looked at (https://tinyurl.com/odiPets2) how the Boston Women’s Workforce Council (BWWC – https://thebwwc.org) has partnered with Boston University to use an MPC to enable companies from the Greater Boston area to collectively compute the sum of their payroll data without revealing their individual contributions. The MPC protocol was produced to benchmark employers and produce a pay equity report covering the Greater Boston area. Introducing MPC technologies to the NDL could enable researchers to conduct similar analyses using UK public sector data.


A further significant consideration for the NDL is the data skills gap.


October-November 2024


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44  |  Page 45  |  Page 46  |  Page 47  |  Page 48  |  Page 49  |  Page 50  |  Page 51  |  Page 52  |  Page 53  |  Page 54  |  Page 55  |  Page 56  |  Page 57  |  Page 58  |  Page 59  |  Page 60