Source-linked AI summary
Artificial Intelligence for the Public Sector: Opportunities and challenges of cross-sector collaboration
Slava Jankin Mikhaylov, Marc Esteve, Averill Campion
TL;DR
Cross-sector collaboration is needed to integrate AI into public-service delivery, but such collaborations face serious management challenges. This paper synthesises existing knowledge and proposes management strategies, identifying facilitative leadership and aligned goals as key factors for collaboration success.
Problem
Cross-sector collaborations face serious management challenges despite their importance for integrating AI and data science into public-sector delivery.
Method
The paper synthesises existing knowledge on managing cross-sector collaborations and proposes recommendations for integrating AI and data science.
Results
Facilitative leadership and alignment of goals and objectives across involved parties are identified as key factors for collaboration success.
Takeaways & Limitations
The proposed strategies imply increased conflict resolution and more inclusive agenda setting in cross-sector collaborations.
Takeaways & Limitations
The paper notes that machine-learning systems may complicate accountability and risk exacerbating bias in technological interfaces.
Abstract
from arXiv · showhide
Public sector organisations are increasingly interested in using data science and artificial intelligence capabilities to deliver policy and generate efficiencies in high uncertainty environments. The long-term success of data science and AI in the public sector relies on effectively embedding it into delivery solutions for policy implementation. However, governments cannot do this integration of AI into public service delivery on their own. The UK Government Industrial Strategy is clear that delivering on the AI grand challenge requires collaboration between universities and public and private sectors. This cross-sectoral collaborative approach is the norm in applied AI centres of excellence around the world. Despite their popularity, cross-sector collaborations entail serious management challenges that hinder their success. In this article we discuss the opportunities and challenges from AI for public sector. Finally, we propose a series of strategies to successfully manage these cross-sectoral collaborations.
Introduction
AI offers substantial opportunities to improve public-service delivery, but realizing them depends on effective collaboration among government, industry, and academia. This study synthesises knowledge on managing cross-sector collaborations and proposes recommendations for integrating AI and data science into public services.
- Challenges: Public-sector AI adoption is hindered by path dependency, information silos, limited resources, weak collaborative culture, and insufficient technical capacity.Successful delivery therefore requires understanding collaboration challenges and success factors, including lessons from other sectors.
- Study contribution: The study synthesises existing knowledge on managing cross-sector collaborations and proposes recommendations for integrating AI and data science into public-service delivery.Existing AI and data-science labs connect government with universities, businesses, and other sectors to combine capabilities, develop solutions, and support public-servant training.
- Challenges: Cross-sector collaborations combine complementary capabilities but involve management complexities, and many fail to achieve satisfactory outcomes.Collaborations range from formal public-private partnerships to informal policy networks and are intended to increase public-service effectiveness and value for money.
Challenges for successful collaboration
Cross-sector AI collaboration is hindered by differences in organisational environments, risk approaches, institutional logics, values, skills, and data capabilities. Accountability for AI decisions, potential misuse, jurisdictional complexity, and bias further complicate collaboration, while mission-oriented projects and knowledge exchange may help bridge differences.
- Organisational environments: Public and private organisations face conflicting accountability environments, creating clashes when aligning partners’ interests.Public organisations answer to service users and the wider public, whereas private organisations respond to shareholders.
- Accountability and bias: AI collaboration raises unresolved accountability questions about machine-learning decisions, including responsibility for adverse impacts and the systems’ responses to complex situations.Without adequate safeguards, new technological interfaces may exacerbate problems of bias.
- Risk and data: Different risk approaches make political risks of government difficult to reconcile with market risks in business organisations.AI data may also be gamed or sabotaged through misleading, destroyed, altered, or injected data during training or operation.
- Institutional logics and values: Competing institutional logics and organisational values complicate agreement on goals and can pit public value against shareholder profit.Public employees are associated with serving the public, while private counterparts seek to further their organisation’s interests, making a shift from “us and them” to “we” necessary.
- Skills and data: Cross-sector AI initiatives face a significant skills gap, with public organisations lacking prerequisite knowledge and skills for effective participation in data-based initiatives.Technical assistance and training are needed, including to personalise disparate public data and operate new systems.
Success factors of collaboration
Successful cross-sector collaboration depends on seven managerial strategies: facilitative leadership, shared objectives, knowledge gathering and sharing, communication, socialising, expertise, and sense-making. Together, these strategies support conflict resolution, inclusive agenda shaping, institutional capacity, unity of purpose, and power balance.
- Facilitative Leadership: Facilitative leadership promotes participation, influence, productive group dynamics, and constructive conflict resolution, especially where incentives and resources are uneven.Leaders should foster respect, positive relationships, candid expression, and broad control rather than impose views hierarchically.
- Shared Objectives: Aligning goals and objectives gives collaboration a shared purpose, guides decision-making, and helps prevent tensions from producing status quo bias.Alliances may combine shared value creation with private value appropriation, while participating organisations can retain distinct objectives.
- Knowledge gathering and sharing: Knowledge gathering and sharing builds institutional and technical capacity through common data standards and the inclusion of frontline and specialist expertise.Frontline workers can identify relevant features and datasets, while network members help align standards with functional responsibilities.
- Communication: A well-defined communication strategy aligns members’ interests and expectations, including their understanding of the opportunities data science and AI offer.Communication is one of the identified managerial strategies for collaboration success.
- Socialising, expertise, and sense-making: Socialisation, expertise, and sense-making complement technical implementation by communicating public value, incorporating alternative problem-solving dimensions, and adapting management to situational understanding.These strategies are identified as key factors in cross-sectoral collaboration success.