Challenge 10.7

Why does this Challenge need to be solved?

There is a perceived rise of distrust in democratic governments around the world. Evidence suggests that the UK public holds an increasingly negative perception of the government’s ability to manage technological innovations – including AI – for public benefit. At the same time, more than 50% of UK adults think that personal data shared with public agencies is not used solely for legitimate reasons. This unfavourable public perception, on the other hand, is seen as major blocker to the wider reuse of public sector data for public benefit.

Scottish Government’s Public Dialogue on the Use of Data has suggested that more citizen participation is needed on a case-by-case basis. This recommendation aligns with the wider aspiration of the Scottish Government to evolve public participation from single initiatives to routine government business.

An ongoing, case-by-case approach to a participation dialogue would allow us to understand and act on the public’s expectations in a nuanced and timely fashion, increasing the probability of getting data-sharing and AI policies right for the Scottish people. However, in times of significant financial pressures, high-quality participation approaches, as catalogued in the Scottish Government Participation Framework, may prove to be too resource intensive to use for specific questions. 

We believe that technology could help us deliver a high-quality, efficient and scalable participation around data sharing and AI use in the Scottish Government.

The Scottish Government currently deploys a range of technologically-enabled participation methods, of which online consultation is the most commonly used in policy development. However, specialists in participation within the Open Government team, emphasise the limitations of online consultations, reflecting a widespread concern that they are too easily dominated by special interest groups.

Online consultations only gather opinion at one point in time and from single respondents, rather than the considered opinion of citizens, which they might arrive at as a group after deliberation. However, deliberative mechanisms that ensure a more informed and considered response from a representative group of participants are resource-intensive to administer whether in person or online.

While new technological solutions can reduce the cost of some stages of some participation processes, it is unclear how to combine these effectively and what the effect is on the quality of the participatory or policy outcomes. It is also important to guard against attacks on such systems by bots or through the spread of misinformation. It is therefore paramount that any proposed technological solution can be rigorously evaluated.

How will we know the Challenge has been solved?

1. Quality methods from the Scottish Government Participation Framework are deployable on an ongoing basis to enable public participation on questions around data, automation, and nuanced policy.

2. Data owners across the public sector are more comfortable with making decisions about data-sharing and automation, reassured that the public have been suitably engaged.

3. An evidence base is widely available for the relative merits of different digital participation tools, helping policy professionals choose a proportionate tool for the public impact of their policy question.

Who are the end users likely to be?

• Open Government Team

• Policy professionals

• Data and AI professionals

• Public engagement and communications professionals

• Social researchers

Has the Challenge Sponsor attempted to solve this problem before?

The Digital Directorate has commissioned premarket research from experts in deliberation technology at the University of Edinburgh and ISWE Foundation – systematically documenting existing digital participation tools. This research has been conducted based on publicly available information only and should be considered as standalone from the CivTech challenge. The market review of existing civic technologies for participation is separate from the CivTech process and will play no part in the CivTech assessment of proposals. Applicants will be expected to address the specific problems set out in the CivTech challenge.

The selected Challenge team will be expected to draw upon the learnings from this research to avoid duplication of existing tools; and ensure uniqueness, suitability and robustness of their proposed solution.  

The research is available here: https://www.gov.scot/isbn/9781836016120

Are there any interdependencies or blockers?

Interdependencies:

• Open Government National Action Plan Commitments - Membership of the Open Government Partnership requires us to deliver concrete commitments to transparency, accountability and participation through National Action Plans. The current plan includes commitments open and ethical use of data and digital technology, and public participation.

• Work of the Scottish Collective Intelligence Community (SCIC) - The SCIC investigates how could institutions and processes be systematically designed for human and computer information sharing and processing. The SCIS aims to enhance the epistemic and decision-making functions of Scottish institutions through innovative applications of Collective Intelligence.

• Ethical Digital Nation, Holistic Ethics, Unlocking The Value of Data and the Scottish AI Strategy – SG is committed to delivering trustworthy, ethical and citizen-centred digital services, AI and data-led activities. Ethics is at the core of what we do.

• Inclusion and Digital Inclusion workstreams – A solution to this Challenge needs to consider, and if possible, address the impact on citizens who cannot or do not wish to engage digitally.

Blockers:

• Tension between techno-solutionism and human-centric or systems approaches to innovation. While we are looking for technological gains in process efficiency, that solution is only expected to enhance human capabilities, keeping real people at the centre of problem-solving and interpretation. We are not expecting technology alone to improve participatory outcomes, but that reducing administrative costs might enable other separate process improvements.

Will a solution need to integrate with any existing systems / equipment?

No.

Any technologies or features the Challenge Sponsor wishes to explore or avoid?

CivTech is tech agnostic. As long as the proposed solution offers the opportunity to solve the Challenge in question, we will consider it.

That said, any proposed solution must be capable of integrating with existing systems as required by the specific Challenge and its Challenge Sponsor, and if appropriate be compatible with current and developing Scottish Government infrastructure.

Much is currently being made of the potential of advanced AI. In truth, just about all the products CivTech has developed over the past few years have AI as part of the tech stack but there is no obligation on your part to go down this route – either with componentry such as machine learning and pattern recognition, or indeed LLMs.

We are looking for the best solution, whatever technology used.

That said, while not limited to the following technologies, the Challenge Sponsor is particularly interested in proposals that explore the following: 

• AI – large language models, conventional Natural Language processing 

• Collective Intelligence – network and agent modelling, complexity, cognitive and behavioural insights, federated learning, sampling theory 

• Socio-technical Systems Thinking 

• Distributed Ledger Technology 

• Transparency and social responsibility

• Open licensing or creative commons of technology within a services business model

What is the commercial opportunity beyond a CivTech contract?

Beyond immediate applications for data and AI use cases, or participation involving the Open Government team, Scottish Government will have extensive need for scalable quality participation to support policy development.

Democratic institutions are under challenge globally, and digital participatory tools are widely considered to be part of counteracting this “backsliding”. So, a solution to this Challenge, to help systematically explore participatory tools, could find application all over the world.

The UK sees widespread, obligatory, use of online consultations, with opportunity widely recognised for higher quality participation methods instead. The UK’s Responsible Technologies Adoption Unit will also have interest in participation solutions for data ethics. A solution developed for Scotland might therefore find ready use expanded to the rest of the UK.

Who are the stakeholders?

Scottish Government staff from the following divisions would be expected to engage and stay apprised of the Applicant’s work;

• Open Government staff across Policy and Communications 

•  Digital Citizen Division / Unlocking Value of Data team 

•  AI Policy and Delivery team

The Scottish Collective Intelligence Community, have been involved and would continue to help to place challenge work in a wider context.

Who’s in the Challenge Sponsor team?

Scottish Government staff from the following divisions;

• Chief Data Office  

• Open Government team

• Unlocking Value of Data team 

What is the policy background to the Challenge?

Scottish Government are a member of the Open Government Partnership, with published commitments to improving our processes for public participation. The Open Government Team responsible for these commitments have expressed interest in wider use of technology coming from this Challenge, for policy issues such as the Just Transition. 

Our Unlocking the Value of Data team have conducted expert and public panel engagement around data sharing, while the AI Policy team have worked with our AI Alliance partners to run public participation around AI Ethics. These engagements have recommended further and more specific engagement around individual use cases for public data.