pith. sign in

arxiv: 2606.13039 · v1 · pith:EE7Z2DSDnew · submitted 2026-06-11 · 💻 cs.CY · cs.AI· cs.HC

Fault Lines: Navigating Ethics and Responsible AI Where National Policy Meets Local Practice in Public Sector Transformation

Sitong Lyu , Shabnam Taghiyeva , Mohit Kukadia , Denis Newman-Griffis This is my paper

Pith reviewed 2026-06-27 05:39 UTC · model grok-4.3

classification 💻 cs.CY cs.AIcs.HC
keywords responsible AIpublic sectorlocal authoritiesnational policyAI governanceSpecial Educational NeedsUK public servicesthematic analysis
0
0 comments X

The pith

Responsible public sector AI requires both national policy adjustments and structural reforms to local institutional capacity, values, and governance.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper examines how UK national AI policies translate into practice at the local authority level, using Special Educational Needs and Disabilities services as a case study to surface tensions. Through thematic analysis of interviews, it identifies five interconnected challenges: shadow AI usage with privacy risks, market-government asymmetry, workforce unreadiness, absent standards for definitions and measurement, and accountability gaps. The authors conclude that responsible AI deployment in public services cannot succeed through national policy alone and must include local-level changes to institutions and oversight.

Core claim

Analysis of 17 interviews reveals five challenges at the national-local interface: shadow usage of AI and data privacy risks, market-government asymmetry in AI provision, insufficient workforce readiness, lack of standardised definitions and measurements, and gaps in human accountability. These intensify in SEND contexts around fairness, oversight, and high-stakes decisions for vulnerable children. The paper concludes that responsible public sector AI therefore requires national policy adjustments together with structural reforms to institutional capacity, values, and governance mechanisms at the local level.

What carries the argument

Thematic analysis of semi-structured interviews identifying barriers and enabling conditions at the national-local interface for responsible AI implementation.

If this is right

  • Strengthening data protection frameworks would reduce shadow AI usage and privacy risks in local authorities.
  • Rebalancing the market-government relationship would address asymmetries in AI tool provision and procurement.
  • Building workforce capacity through training would improve readiness for responsible AI use.
  • Establishing standardised definitions and metrics would enable consistent measurement of responsible AI outcomes.
  • Reinforcing human accountability mechanisms would better handle high-stakes decisions in areas like SEND.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Without dedicated local funding or support, national AI strategies may remain aspirational rather than operational in resource-constrained authorities.
  • The emphasis on high-stakes domains like SEND implies that sector-specific rules may be needed beyond general principles for vulnerable groups.
  • Similar interface problems could appear in other devolved public services such as social care or housing, suggesting the pattern is not unique to education.

Load-bearing premise

The accounts from 17 interviews with a convenience sample of policymakers, practitioners, and third-sector professionals are sufficient to identify the main fault lines and reflect actual local practice without major selection or reporting bias.

What would settle it

A survey or audit of a larger, representative set of local authorities demonstrating consistent successful responsible AI deployment aligned with national policy and without the listed challenges would undermine the claim that structural local reforms are required.

read the original abstract

The UK government has adopted a pro-AI stance to help transform public service delivery in the face of severe financial pressures, but the path to translate this vision into responsible AI practice remains ill-defined. While UK policy is often set at the national level, local authorities are responsible for most public service delivery, and the rapid advance of AI-first narratives in the public sector is exposing fault lines in knowledge and practice at this national-local interface. This paper examines how responsible AI is interpreted and implemented at the interface between the UK's central government and local authorities, taking the high-stakes area of Special Educational Needs and Disabilities (SEND) as a case study. We present a thematic analysis of 17 semi-structured interviews with policymakers, practitioners, and third-sector professionals to identify barriers and enabling conditions for responsible AI where national policy meets local practice. We identify five interconnected challenges facing local authorities: shadow usage of AI and data privacy risks, market-government asymmetry in AI provision, insufficient workforce readiness, a lack of standardised definitions and measurements, and gaps in human accountability. For each, participants proposed actionable steps, from strengthening data protection frameworks and rebalancing the market-government relationship to enhancing workforce capacity. Our examination of SEND brings these challenges into sharper focus, showing how high-stakes decisions affecting vulnerable children and families intensify tensions around accountability, fairness, and human oversight, exposing the limits of a principle-based regulatory approach. We argue that responsible public sector AI requires both national policy adjustments and structural reforms to institutional capacity, values, and governance mechanisms at the local level.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 0 minor

Summary. The paper examines how responsible AI is interpreted and implemented at the UK national-local government interface, using Special Educational Needs and Disabilities (SEND) services as a case study. It presents a thematic analysis of 17 semi-structured interviews with policymakers, practitioners, and third-sector professionals, identifying five interconnected challenges (shadow AI usage and privacy risks, market-government asymmetry, workforce readiness gaps, lack of standardised definitions/measurements, and human accountability gaps). The authors argue that responsible public sector AI requires both national policy adjustments and structural reforms to local institutional capacity, values, and governance mechanisms, with participant-proposed actionable steps.

Significance. If the thematic analysis is shown to be robust and representative, the work would offer timely empirical insight into practical barriers at the national-local interface for AI ethics in public services, particularly in high-stakes domains like SEND. It draws directly from practitioner accounts to propose concrete steps and highlights limits of principle-based regulation. However, the absence of methodological transparency currently limits its contribution to the evidence base on public sector AI governance.

major comments (1)
  1. [Methods] The description of the thematic analysis (mentioned in the abstract and presumably detailed in the Methods section) provides no information on the coding process, saturation criteria, participant selection beyond a convenience sample of 17 interviewees, or study limitations. This is load-bearing for the central claim, as the five challenges and the prescription for national adjustments plus local structural reforms rest on these themes accurately capturing primary fault lines without material selection or reporting bias.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their detailed and constructive review. The feedback on methodological transparency is well-taken and directly addresses a key requirement for the credibility of our qualitative findings. We address the single major comment below and will incorporate revisions to strengthen the manuscript.

read point-by-point responses

  1. Referee: [Methods] The description of the thematic analysis (mentioned in the abstract and presumably detailed in the Methods section) provides no information on the coding process, saturation criteria, participant selection beyond a convenience sample of 17 interviewees, or study limitations. This is load-bearing for the central claim, as the five challenges and the prescription for national adjustments plus local structural reforms rest on these themes accurately capturing primary fault lines without material selection or reporting bias.

    Authors: We agree that the current Methods section does not provide sufficient detail on the thematic analysis procedures, and this omission limits the ability to evaluate the robustness of the five identified challenges. In the revised version we will expand the Methods section to describe: (1) the coding process, including the use of inductive thematic analysis with initial open coding followed by axial coding to identify interconnections; (2) saturation criteria, noting that thematic saturation was assessed iteratively and considered reached when no new themes emerged in the final three interviews; (3) participant selection, clarifying that while convenience sampling was used for accessibility, it was combined with purposive targeting of specific roles (policymakers, practitioners, and third-sector experts in SEND and digital transformation) to ensure relevance; and (4) study limitations, including the modest sample size, potential for self-selection bias, geographic concentration in England, and the interpretive nature of the analysis. These additions will be placed in a dedicated subsection and will explicitly link back to the validity of the reported fault lines and participant-proposed steps. We do not claim the sample is statistically representative; the revisions will make this explicit. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical thematic analysis grounded in primary interview data

full rationale

The paper contains no equations, derivations, fitted parameters, or predictive claims that reduce to inputs by construction. Its central argument—that responsible public sector AI requires national policy adjustments plus local structural reforms—is presented as emerging directly from thematic analysis of 17 semi-structured interviews. No self-citation load-bearing steps, uniqueness theorems, or ansatzes are invoked to justify the findings; the analysis is self-contained as a qualitative case study of the UK national-local interface in SEND services. This matches the default expectation for non-circular empirical work.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The paper is a qualitative policy study with no mathematical content. It rests on standard social-science assumptions about the value of interview data rather than introducing free parameters, new entities, or ad-hoc axioms beyond the domain norm that thematic analysis of practitioner accounts can surface actionable governance insights.

axioms (1)
  • domain assumption Thematic analysis of semi-structured interviews with a small purposive sample can reliably surface the main barriers and enabling conditions for responsible AI at the national-local interface.
    Invoked implicitly when the authors move from 17 interviews to five general challenges and policy recommendations.

pith-pipeline@v0.9.1-grok · 5828 in / 1206 out tokens · 31277 ms · 2026-06-27T05:39:29.238243+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

3 extracted references

  1. [1]

    Challenges of responsible AI in practice: scoping review and recommended actions

    Malak Sadek, Emma Kallina, Thomas Bohné, Céline Mougenot, Rafael A. Calvo, and Stephen Cave. 2025. “Challenges of responsible AI in practice: scoping review and recommended actions”. AI & Society 40 (Jan. 2025), 199–215. https://doi.org/10.1007/s00146-024-01880-9 [9] Petar Radanliev, Omar Santos, Alistair Brandon-Jones, and Adam Joinson. 2024. Ethics and ...

    doi:10.1007/s00146-024-01880-9 2025

  2. [2]

    AI Ethics

    Hanne Hirvonen. 2024. Just accountability structures – a way to promote the safe use of automated decision-making in the public sector. AI & Society 39 (2024), 155–167. https://doi.org/10.1007/s00146-023-01731-z [24] Viviana Bastidas, Kwadwo Oti-Sarpong, and Jennifer Schooling. 2025. Responsible digitalisation and AI use for local authorities: Report of f...

    doi:10.1007/s00146-023-01731-z 2024

  3. [3]

    Thomas Vogl. 2021. Artificial intelligence in local government: Enabling artificial intelligence for good governance in UK local authorities. Report. University of Oxford, Oxford, UK (April 2021). Retrieved from https://ora.ox.ac.uk/objects/uuid:60fad1d4-b297-4070-81ce-36c8d18b4dd5/files/s3n203z83r [39] Tan Yigitcanlar, Duzgun Agdas, and Kenan Degirmenci....

    doi:10.1007/s00146-022-01450-x 2021