Dr. Smrite Goudhaman smritegoudhaman.com All Research
Peer-Reviewed Research · CERE 2026 · Paper ID 41

Designing Trustworthy AI — Human Oversight Lessons from India

A two-phase longitudinal study across 350 frontline workers in a live operational environment — measuring what happens when AI governance moves from policy document to platform architecture.

Authors
Goudhaman, Dahiya & Kumar (2026)
Institution
Golden Gate University, San Francisco, USA
Deployment Site
Toscano Restaurant Chain, India · SafetyCulture Platform
Study Period
2024 – 2025 · n = 350 frontline workers
Presented At
IIM Indore · IIM Kashipur · IIT Delhi · AAAI Singapore · AITC Nuremberg · UNESCO Paris · UN Geneva
CERE 2026 · IIM Indore MERC 2026 · IIM Kashipur AI Impact Summit · IIT Delhi AAAI-26 · Singapore AITC 2026 · Nuremberg IASEAI · UNESCO Paris UN Global AI Governance · Geneva
350
Frontline workers across two phases
86.82%
Completion rate at scale (above 85% threshold)
+4.26
Assessment score improvement at scale
−57%
Reduction in time-on-task (learning efficiency gain)

Why Frontline Workers Are at the Centre of the AI Governance Question

The IMF estimates that 40% of global jobs are now exposed to AI disruption — yet only 26% of jobs in low-income countries face AI exposure, compared to 60% in advanced economies. This asymmetry does not mean developing economies are protected. It means they have less capacity to benefit and fewer governance frameworks equipped to respond.

Frontline workers in high-churn service sectors — hospitality, retail, food service — are simultaneously the most exposed and the least supported by reskilling infrastructure. They are also the population most often excluded from the design of AI systems meant to serve them.

This research was designed precisely for them.

40%

of global jobs exposed to AI disruption

IMF · Gen-AI and the Future of Work (Jan 2024)

92M

jobs projected to be displaced globally by 2030

Yet 170M new roles also projected — net gain depends entirely on governance and reskilling investment.

WEF Future of Jobs Report 2025

26%

AI exposure in low-income countries vs. 60% in advanced economies

IMF (2024) — same source

"Technology scales fast. Trust scales slowly. And this asymmetry is the central governance challenge of our era."

— Goudhaman, Dahiya & Kumar (2026) · CERE 2026 IIM Indore · Paper ID 41

What Frontline Deployment Exposes That Policy Frameworks Miss

Most AI governance frameworks are written at the level of principle. They articulate values — fairness, transparency, accountability — but leave a critical gap between declaration and deployment. This research was designed to close that gap.

Two years of platform-generated behavioural data from a real operational environment gave us something that policy documents rarely have: objective, longitudinal evidence about what actually happens when AI enters a frontline workplace.

No self-report data. No survey instruments. Behavioural trace only — completion records, assessment scores, time-on-task per module. This methodological choice was deliberate: to address common method bias (Podsakoff et al., 2003) and to ground every finding in what workers actually did, not what they said they would do.

  • Fear and resistance to AI adoption Including algorithm aversion risk (Dietvorst et al., 2015) — workers abandoning AI after a single perceived failure
  • Low engagement and training fatigue When volume scaled 15× without governance architecture, engagement risk increased sharply
  • Accountability gaps in live deployment Who is responsible when an AI-guided decision goes wrong? This study built the answer into the architecture.
  • Digital exclusion of low-literacy workers Frontline workforces span education levels, languages, and digital literacy — most AI systems are not designed for this reality
  • Governance treated as compliance burden When governance is a checklist rather than architecture, it disappears the moment operational pressure increases

Two-Phase Longitudinal Study: Toscano Restaurant Chain, India

The study was designed to test what happens when AI governance shifts from a controlled doctoral pilot to a live operational environment at 2.5× scale — with reduced direct supervision and AI governance as the primary accountability layer.

2024

Phase 1 · Controlled Doctoral Pilot · n = 100

  • 100 frontline workers — controlled environment
  • Direct managerial oversight throughout
  • Supervisory-led deployment
  • High-contact governance model
  • Baseline data established across all metrics
2025

Phase 2 · Live Operational Scale · n = 250

  • 250 frontline workers — live workflows
  • 15× training volume increase
  • Reduced direct supervision
  • AI governance as primary accountability layer
  • Same platform configuration — no variable changes
Reminders

Timed completion notifications. No decision-making authority. No escalation capability.

Pacing Cues

Adaptive content sequencing based on learner progress. Algorithm adapts; human reviews outcomes.

Reinforcement Nudges

Post-assessment feedback supporting correct responses. Supportive only — no punitive signals.

Constrained agency design follows Amershi et al. (2019) · doi.org/10.1145/3290605.3300233

Human Oversight Built Into the Structure — Not Added as Audit

The study introduced a new theoretical concept: Constrained Agency — AI decision-making authority deliberately restricted to specific, low-stakes task-level functions. All higher-order decisions remain exclusively with human managers.

Human Review

All AI outputs subject to manager review before action

Role Bounding

AI restricted to 3 task functions only — no scope creep

Escalation

Exception-handling stays exclusively with human managers

Audit Trails

Platform logs serve as accountability evidence — objective behavioural trace

Trust by Design

No algorithm aversion detected in either phase — governance was the trust anchor

"Constrained Agency is not a limitation on AI. It is the condition under which AI earns the right to be trusted."

— Theoretical contribution, Goudhaman, Dahiya & Kumar (2026) · Building on Hemmer et al. (2025) and Amershi et al. (2019)

What the Evidence Shows — All Figures From the Published Paper

Completion Rate 2024
100%
Completion Rate 2025
86.82%
Assessment Score 2024
79.12
Assessment Score 2025
83.38
Theoretical Contribution · The Inverse Learning Paradox

Scores rose +4.26 points. Time-on-task fell 57%. Both happened simultaneously.

When AI is constrained and governed, speed and learning depth are not in conflict — they are complementary. Workers who spent less time per module scored higher on assessments. This finding challenges the assumption that reduced time-on-task signals disengagement. In a governed AI environment, it signals learning fluency and consolidation. The Inverse Learning Paradox is a new theoretical contribution to the AI governance and learning science literature.

Metric & Hypothesis Pilot 2024 (n=100) Scale 2025 (n=250) Change & Interpretation
Completion Rate H1 — AI Adoption 100% 86.82% ✓ H1 SUPPORTED

−13.18% — operational normalisation, not disengagement. Above the 85% institutional threshold. AI micro-agents sustained adoption despite 15× volume increase.

Assessment Score H2 — Learning Quality 79.12 pts 83.38 pts ✓ H2 SUPPORTED

+4.26 pts — deeper learning at scale despite reduced supervision. AI reinforcement cues drove consolidation of correct responses.

Time-on-Task H3 — Learning Efficiency 10.4 min 5.98 min ✓ H3 SUPPORTED

−4.42 min (−57%) — learning fluency and efficiency gain. Inverse correlation with quality = Inverse Learning Paradox. The Paradox reframes efficiency as a proxy for mastery, not disengagement.

Algorithm Aversion H4 — Trust in AI None detected None detected ✓ H4 SUPPORTED

No aversion across either phase. Governance structures sustained continuous AI engagement. Managerial oversight was the trust anchor — not the AI's capability alone.

All figures: SafetyCulture platform behavioural logs 2024–2025 · Toscano Restaurant Chain, India · Goudhaman, Dahiya & Kumar (2026) CERE IIM Indore Paper ID 41

What Developing Economies Contribute to AI Governance

"Developing economies are not 'behind.' They are designing adaptive, resilient, human-centred AI governance under real-world constraints — and the world must learn from them."

Workforce Inclusion

AI designed for workers across education levels, languages, and digital literacy — not just skilled users. This is not an accommodation. It is a design standard that raises the floor for everyone.

Multilingual Realities

High-churn, diverse, multilingual frontline workforces as a design constraint — not an edge case. The governance framework must hold across this diversity.

Low-Resource Innovation

Trustworthy AI built under real budget and infrastructure pressure — producing governance models that are replicable globally, not just in well-resourced environments.

Adaptive Governance

Governance that bends to operational context rather than imposing uniform top-down templates. The pilot-to-scale playbook is replicable in healthcare, retail, and manufacturing.

Human-First Design

Workers as active participants in trust-building — not passive subjects of AI transformation. This is not a soft principle. It is what the behavioural data shows drives adoption.

Stress-Tested Models

AI deployment stress-tested by real-world complexity, churn, and 15× scale increase — producing evidence about what governance architectures actually hold under pressure.

Five Principles for Trustworthy AI Deployment at Scale

Each recommendation is grounded in a specific finding from this longitudinal study — not in abstract principle, but in what 350 frontline workers' behavioural data showed across two years of live deployment.

01

Mandate Human Oversight Architecturally

Embed review, escalation, and role bounding into deployment design from the beginning — not as audit, but as structure. Governance that is architectural cannot be switched off under operational pressure. Governance that lives only in a policy document will be.

Finding: governance architecture prevented algorithm aversion across both phases. Zero aversion detected.
02

Measure Trust and Efficiency — Not Just Completion

Completion rates alone are insufficient signals of AI adoption quality. Policymakers and organisations must track assessment quality, time-on-task, and aversion indicators in parallel. The Inverse Learning Paradox demonstrates why: a falling metric (time-on-task) masked a rising outcome (learning quality).

Finding: completion fell 13.18% while quality rose +4.26 pts — the opposite of what a completion-only lens would suggest.
03

Include Frontline Workers in Governance Design

Frontline learners are stakeholders in governance, not end-users of a system designed without them. Agency over pacing, escalation rights, and visible accountability structures sustained engagement without producing aversion. This is structural trust-building — not soft policy.

Finding: no aversion in either phase, despite 15× training volume — workers remained active participants throughout.
04

Apply Constrained Agency as a Design Principle

Restrict AI authority to specific, low-stakes task-level functions. All higher-order decisions — assignment, escalation, exceptions — must remain human-held. Role bounding is the operational difference between AI as assistive scaffolding and AI as threatening automation.

Finding: three constrained functions (reminders, pacing, reinforcement) sustained adoption at 86.82% across 250 workers.
05

Treat AI Governance as Operational Practice — Not Policy Document

Governance must live in platform architecture, manager training, data infrastructure, and review cycles. This study's pilot-to-scale playbook is replicable across healthcare, retail, and manufacturing — anywhere frontline workers intersect with AI in high-stakes, low-resource environments.

Finding: SafetyCulture platform logs functioned as real-time governance audit trail — making accountability visible at scale.
86.82%
Completion at scale — above the 85% threshold
+4.26
Assessment gain — quality improved as scale increased
−57%
Time-on-task — accelerated learning consolidation

"The future of AI governance will not be decided only in laboratories or policy rooms. It will be decided where people learn, work, struggle, adapt — and choose to trust."

Goudhaman, Dahiya & Kumar (2026) · CERE 2026 · IIM Indore · Paper ID 41

The four foundations of trustworthy AI

🤝
Trust
Earned through consistent behaviour, not declared through communication
🌱
Dignity
Every learner treated as a full person — not a data point to be optimised
🔍
Transparency
Visible reasoning at every step — not post-hoc explanation of opaque decisions
💬
Dialogue
Ongoing conversation between AI systems, organisations, and the people they serve

Trustworthy AI is not built through automation alone.

How to Cite This Work

Primary Citation

Goudhaman, S., Dahiya, R., & Kumar, A. (2026). Designing Trustworthy AI: Human Oversight Lessons from India for Developing Economies. Proceedings of the 16th International Conference on Excellence in Research and Education (CERE 2026). IIM Indore. Paper ID 41.

  • Amershi, S. et al. (2019). Guidelines for Human-AI Interaction. CHI 2019. doi.org/10.1145/3290605.3300233
  • Dietvorst, B. J., Simmons, J. P., & Massey, C. (2015). Algorithm aversion: People erroneously avoid algorithms after seeing them err. Journal of Experimental Psychology: General, 144(1), 114–126. doi.org/10.1037/xge0000033
  • Fügener, A. et al. (2019). Will humans-in-the-loop become borgs? Merits and pitfalls of human and machine collaboration. Journal of Information Systems Research. doi.org/10.2139/ssrn.3368813
  • Hemmer, P. et al. (2025). Human-AI collaboration in decision-making. European Journal of Information Systems. doi.org/10.1080/0960085X.2025.2475962
  • IMF (2024). Gen-AI: Artificial Intelligence and the Future of Work. IMF Staff Discussion Note SDN/2024/001. imf.org
  • Podsakoff, P. M. et al. (2003). Common method biases in behavioral research. Journal of Applied Psychology, 88(5), 879–903. doi.org/10.1037/0021-9010.88.5.879
  • Venkatesh, V. et al. (2003). User acceptance of information technology: Toward a unified view. MIS Quarterly, 27(3), 425–478.
  • WEF (2025). The Future of Jobs Report 2025. World Economic Forum. weforum.org
↓ Download Research Summary Request Full Paper ← Back to the Website