Focus
Social sustainability
Worker protection, rights, and dignity across global supply chains, the side of sustainability hardest to measure.
Section i
About the research
A study of how Artificial Intelligence is being used in global supply chains to protect workers and uphold their rights, with focus on the social side of sustainability that is easiest to overlook.
Scope
Full adoption lifecycle
From the decision to explore AI, through implementation, to the evaluation of its real impact on the ground.
Method
Practitioner interviews
Semi-structured 45 to 60 minute conversations with practitioners whose work shapes how these tools are used.
This study looks at how Artificial Intelligence (AI) is being used in global supply chains to protect workers and uphold their rights. While environmental sustainability is often well measured, the social side, things like fair wages, safe working conditions, and freedom from exploitation, is harder to track and easier to overlook.
We are exploring the full journey of AI adoption: how organisations decide to bring these tools in, how they put them into practice, and what difference they actually make for workers on the ground. Your professional experience will help us understand what is working, what is not, and where the real opportunities and risks lie.
“Your professional experience is invaluable. There are no right or wrong answers. What matters is your honest perspective on how AI truly operates in practice.”
Section ii
What our conversation will be about
A relaxed, open discussion about your professional experience with AI in supply chains. There are no right or wrong answers, we are interested in your perspective and what you have learned in practice.
We will explore three broad areas, guided by your experience so we can spend more time on the parts most relevant to your role.
Area i.
How AI is being adopted
What shapes the decision to bring AI in.
What is shaping decisions to bring these tools into supply chain work, and what is helping or holding back wider use across the sector.
Area ii.
How AI is being used day to day
The practical role in everyday work.
The role these tools play in managing supply chains, including what they do well and where people still add the most value.
Area iii.
What it means for the people in the supply chain
The wider effects on the ground.
How AI is changing the way companies understand and respond to what is happening on the ground, and the wider effects this is having on workers.
The conversation usually takes 45 to 60 minutes. For a normative reference on worker rights, see the ETI Base Code in the Appendix.
Section iii
The interview itself
The conversation takes place online via Microsoft Teams at a time that suits you, and lasts around 45 to 60 minutes in total. It is a one to one discussion, just you and the researcher, with no other participants.
Here is roughly how the time is spent.
i.
Welcome & introductions
A brief welcome, a quick recap of the study, and a chance for you to ask any questions before we begin.
First 5 min
ii.
The main conversation
Guided by a short set of open questions and shaped by your experience. There is no fixed script, so we can spend more time on the areas most relevant to your role.
35 to 45 min
iii.
Close & next steps
A chance to add anything we have not covered, share any final thoughts, and discuss next steps, including how and when you will hear about the findings.
Final 5 to 10 min
Recording
With your permission, the interview will be audio recorded so that we can focus on the conversation rather than note taking. The recording is used only to prepare a written transcript, and the original audio is deleted once that is complete.
You are in control
You can pause, stop, or skip any question at any point, or end the interview early without giving a reason. If you would like to see the questions in advance, we are happy to share them ahead of time.
After the interview
You are welcome to review your transcript if you would like to add, clarify, or remove anything before it is used in the analysis.
Section iv
How your contribution is cared for
Your time and trust matter, and we take care to handle what you share with the same respect.
Your identity stays private.
You will be given a pseudonym, and your name, organisation, and any other identifying details will be removed from the transcript before analysis. Nothing you say will be linked back to you in the thesis, in any publication, or in any presentation of the findings.
Your information is held securely.
Recordings, transcripts, and contact details are stored on Cardiff University's secure servers, accessible only to the researcher and the supervisory team. The original audio is deleted once the transcript is complete.
You stay in control.
Your participation is entirely voluntary. You can withdraw at any point before, during, or after the interview without giving a reason. If you withdraw, any information you have shared can be removed, unless it has already been anonymised and combined into the wider analysis.
Quotes are used carefully.
Short, anonymised excerpts from the interview may appear in the thesis or published work to illustrate a point. These will only ever be attributed by general role and sector, never by name, organisation, or any detail that could identify you.
You will see what comes of it.
Once the analysis is complete, all participants will receive a short summary of the findings, along with an invitation to an online seminar where the results will be shared.
Ethics approval
The study has been reviewed and approved by the Cardiff Business School Research Ethics Committee.
Reference · CARBS SREC 3132
Questions or concerns
Researcher: Zhe Li
liz142@cardiff.ac.uk
Supervisor: Dr Maryam Lotfi
lotfim@cardiff.ac.uk
Appendix
Supplementary references
Optional background for readers wishing to examine the normative lens and the theoretical framework that inform our questions.
A. Normative Lens
The ETI Base Code: nine principles of labour practice
View
A. Normative Lens
The ETI Base Code: nine principles of labour practice
The Ethical Trading Initiative (ETI) Base Code is an internationally recognised code of labour practice founded on the conventions of the International Labour Organisation. It consolidates the core labour standards underpinning most due-diligence legislation and buyer codes of conduct. We use it as the normative lens through which AI's impact on specific worker rights is evaluated.
Clause i.
Employment is freely chosen
No forced, bonded or involuntary labour. Workers are free to leave after reasonable notice; no deposits or identity papers are required.
Clause ii.
Freedom of association
Workers have the right to join or form trade unions and to bargain collectively; employers adopt an open attitude toward union activities.
Clause iii.
Working conditions are safe and hygienic
A safe environment shall be provided; adequate steps shall be taken to prevent accident and injury; workers receive regular health and safety training.
Clause iv.
Child labour shall not be used
No new recruitment of child labour. Policies shall support the transition of any child found performing such labour to quality education.
Clause v.
Living wages are paid
Wages meet at minimum national legal or industry benchmark standards, and should be enough to meet basic needs and provide some discretionary income.
Clause vi.
Working hours are not excessive
Working hours shall not exceed 48 per week; overtime shall be voluntary and compensated at a premium rate; totals shall not exceed 60 hours in any 7-day period save exceptional circumstances.
Clause vii.
No discrimination is practised
No discrimination in hiring, compensation, training, promotion, termination, or retirement on any protected ground.
Clause viii.
Regular employment is provided
Work is performed on the basis of a recognised employment relationship; obligations shall not be avoided through labour-only contracting or excessive fixed-term contracts.
Clause ix.
No harsh or inhumane treatment
Physical abuse or discipline, the threat of physical abuse, sexual or other harassment, and verbal abuse or other forms of intimidation shall be prohibited.
Source: Ethical Trading Initiative. These provisions constitute minimum, not maximum, standards.
B. Theoretical Framework
The three-stage lifecycle of AI adoption for social sustainability
View
B. Theoretical Framework
The three-stage lifecycle of AI adoption for social sustainability
Our earlier literature review identified a three stage lifecycle in how organisations adopt AI for social sustainability. Interview questions map to these stages; the framework itself remains a hypothesis that this empirical study tests and refines.
Stage i.
Pre Adoption
Epistemic readiness
Organisations must resolve technical and ethical uncertainties before committing resources. Absorptive capacity functions as the gateway mechanism.
P1 · Direct. Higher absorptive capacity → progression from awareness to active evaluation of AI for SCSS.
P2 · Mediation. External regulatory pressure → AI adoption, mediated by data-driven organisational culture.
Stage ii.
Adoption
Decision & implementation
Diffusion-of-Innovation theory structures persuasion, decision and implementation. The critical choice is the data-sourcing architecture of the AI itself.
P3 · Institutional trust. Embedded, firm-like AI structures → higher worker trust and lower post-adoption resistance.
P4 · Inverted-U. Moderate monitoring protects; excessive surveillance erodes autonomy and trust.
Stage iii.
Post Adoption
Ethical governance
AI systems are mutable infrastructure. Concept drift erodes performance; governance requires algorithmic transparency, outcome measurement and participatory validation.
P5 · Moderation. XAI moderates the relationship between AI longevity and sustained SCSS performance.
P6 · Compounding. The digital sustainability divide widens at each successive lifecycle stage.
C. Methodology Note
Design, participants and analysis
View
C. Methodology Note
Design, participants and analysis
Design
Qualitative exploratory study using semi structured interviews via Microsoft Teams; audio recorded, transcribed and analysed thematically.
Participants
Professionals in supply chain management, procurement, sustainability or ethical trade with experience of AI tooling or social compliance processes.
Analysis
Thematic analysis: deductive coding against the ETI Base Code and the TOE/DOI framework, supplemented by inductive codes from emergent practitioner themes.