ICSD ESG 博客

本博客專欄旨在讓會員分享關於ESG的想法和經驗。這是一個讓我們相互學習的絕佳平臺。

每篇文章應約為700字（中文或英文）。請發送電子郵件至info@icsd-global.org以提交你的文章。

此ESG博客內的所有文章均為作者個人觀點，並不反映ICSD的立場或政策。本會不對內容的準確性或完整性承擔任何責任。

Fair to bank customers in U.K.: Say NO to biased surveys

Fair to bank customers in U.K.: Say NO to biased surveys

Our UK ESG Watch spotted a regulatory requirement which forces the banks in the U.K. to disclose the quality survey results done by an independent agent in their promotion pamphlets!

The UK's Competition and Markets Authority requires banks to show the survey results to the customers in their promotion materials, this is fair to customers but might be "harsh" to banks. For real-life examples, please see the photos of HSBC UK promotion pamphlets. This requirement applies to both the personal current account providers and the business current account providers.

If you want to know more, please see the survey agent's websites:

Personal banking:

https://www.ipsos.com/en-uk/personal-banking-service-quality

Business banking:

https://www.ipsos.com/en-uk/business-banking-service-quality

Transparency is always a key for enhancing social equity and corporate governance.

Governing Intelligence: Integrating AI Ethics and Data Governance into the UK’s Digital ESG Framework

Governing Intelligence: Integrating AI Ethics and Data Governance into the UK’s Digital ESG Framework

Janet Ng, U.K. ESG Advocate

ymjanetng@gmail.com

In the UK’s rapidly evolving digital economy, artificial intelligence (AI) is transforming business, public administration, and everyday decision‑making. Yet this technological revolution brings not only opportunity, but also profound ethical and governance challenges. Bias, opacity, and misuse of data threaten the very social trust upon which sustainable business depends (House of Lords, 2018; ICO, 2023).

AI governance and data ethics must be integrated into ESG frameworks as core measures of corporate responsibility — not as optional ethics add‑ons. In other words, ethical AI is the new frontier of ESG performance. Only by embedding transparent, fair, and accountable AI systems can the UK create both technological innovation and social legitimacy in the digital age (Leslie, 2019; DSIT, 2024).

Ethical AI Is an ESG Issue

Traditional ESG reporting has focused on carbon footprints, diversity metrics, and board oversight. However, as algorithms increasingly shape how resources, jobs, and information are distributed, AI has become a material governance and social‑impact issue (Leslie, 2019; Floridi, et al., 2018). The rapid growth of AI infrastructure — from energy intensive data centres to the accelerated hardware cycles of machine learning systems — has also considerable environmental consequences (OECD, 2024; DSIT, 2024).

Environment

High computational demand increases carbon emissions, while frequent hardware upgrades contribute to global e‑waste (Floridi, et al., 2018; OECD, 2024). Integrating AI’s environmental footprint into ESG reporting — through indicators such as data‑centre energy intensity, use of renewable energy, and e‑waste recovery — would enable organisations to assess both the ecological and ethical sustainability of their digital operations (CDEI, 2021; DSIT, 2024). In doing so, businesses would complete the E + S + G equation, ensuring technological progress aligns with planetary boundaries as well as human values (Floridi, et al., 2018; Leslie, 2019).

Social

Ethical data is measurable social capital. Transparent data practices sustain public trust and therefore constitute a quantifiable ESG asset (CDEI, 2021; ICO, 2023). If governance is the foundation of ESG, data ethics is its social currency. Organisations now hold unprecedented power over personal and behavioural data. Misuse of that data damages social legitimacy and undermines the “S” of ESG — trust and fairness in stakeholder relations (ICO, 2023; CDEI, 2021). The UK’s Centre for Data Ethics and Innovation (CDEI) defines “trustworthy innovation” as decision‑making that respects privacy, consent, and fairness across the data lifecycle (CDEI, 2021). This transforms data management from an operational task into a moral and social responsibility (Leslie, 2019; Floridi, et al., 2018). Dr Leslie of The Alan Turing Institute (2019) emphasises that responsible data design — using representative, timely, and bias‑checked datasets — is the first ethical barrier against social harm. When companies treat data ethics as integral to ESG, they strengthen social capital: consumer loyalty, employee confidence, and investor assurance (CDEI, 2021; FCA, 2023; Leslie, 2019).

In practice, the UK’s Information Commissioner’s Office (ICO) (2023) requires Data Protection Impact Assessments (DPIAs) for high‑risk AI projects, showing that data transparency is not just good ethics but regulatory ESG compliance. These audits mirror social impact assessments, demonstrating how responsible data governance is now ESG governance (ICO, 2023; DSIT, 2024).

Governance

Responsible AI is a governance imperative. Businesses that ignore AI ethics risk regulatory sanctions, reputational damage, and the erosion of stakeholder trust — core ESG risks (DSIT, 2024; OECD, 2024). The UK’s AI Regulation White Paper (OECD, 2024) affirms that AI must be developed within a governance framework of safety, transparency, fairness, accountability, and contestability — principles that align directly with ESG values. Fairness addresses discrimination; accountability ensures responsible oversight; transparency builds stakeholder trust. These are not only “technical safeguards” — they are corporate governance metrics that investors and regulators can measure and audit (FCA, 2023; DSIT, 2024).

As Dr Leslie (2019) argues, AI ethics must translate values into governance structures. His FAST framework — Fairness, Accountability, Sustainability, and Transparency — maps perfectly onto the S (social) and G (governance) pillars of ESG. Ethical AI, therefore, isn’t just about compliance; it’s about corporate integrity and long‑term sustainability (Leslie, 2019; Floridi, et al., 2018).

Sustainability Through Human‑Centred AI

Ethical AI is social sustainability in action. By ensuring AI systems are transparent, accountable, and inclusive, organisations operationalise the “S” and “G” pillars together (Leslie, 2019; DSIT, 2024). AI’s sustainability implications go beyond energy efficiency. As algorithmic systems begin to shape climate forecasting, healthcare access, and resource allocation, ethical governance becomes central to social sustainability (Floridi, et al., 2018; OECD, 2024). Unchecked bias or discrimination in AI can deepen inequality — contradicting not only ESG ethics but also the UN Sustainable Development Goals (e.g., SDG 16)  (OECD, 2024).

The UK’s AI Report “Ready, Willing and Able?” warns that without transparent oversight, AI risks eroding human dignity and public confidence (House of Lords, 2018). Conversely, when aligned with the Dr Leslie of Alan Turing Institute’s FAST principles and the OECD’s AI recommendations for fairness and accountability (OECD, 2024; Leslie, 2019), AI becomes a force for inclusive, resilient governance. Thus, sustainability in the digital era is not just “green”; it’s ethical and human‑centred. Businesses that ensure AI decisions are explainable, fair, and auditable contribute directly to the longevity and resilience of democratic institutions (Floridi, et al., 2018; CDEI, 2021).

The Public Sector as a Model for Digital ESG

Public‑sector ethical AI frameworks can serve as a governance blueprint for business ESG reporting (Leslie, 2019; FCA, 2023). The UK government and the Alan Turing Institute have established Process‑Based Governance frameworks (PBG) to make algorithmic decisions accountable in public services (Leslie, 2019). These frameworks require documentation such as Dataset Factsheets, Model Cards, and Stakeholder Impact Assessments, ensuring decision traceability. This method mirrors ESG auditing — where organisations record how they identify, mitigate, and disclose risks (FCA, 2023). If extended to the private sector, this “algorithmic audit trails” could form part of future ESG disclosures under the Corporate Sustainability Reporting Directive (CSRD) (DSIT, 2024; CDEI, 2021). By demonstrating accountability‑by‑design, the UK public sector sets a precedent for corporate governance reform. Financial regulators, such as the Financial Conduct Authority (FCA), are already considering integrating AI ethics and data governance metrics into Sustainability Disclosure Requirements (SDR) and investment labels (FCA, 2023).

Conclusion: Toward a Human‑Centred Digital ESG

AI governance and data ethics are the new foundations of ESG integrity — and the measure of whether digital progress in the UK truly serves society (Floridi, et al., 2018; Leslie, 2019). AI’s economic potential will mean little without ethical legitimacy (House of Lords, 2018). As Dr Leslie (2019) puts it, “AI is not morally autonomous; governance must remain a human responsibility (Leslie, 2019).” If ESG once focused on how businesses manage resources and people, the next phase — Digital ESG— will define how they govern intelligence itself. Organisations that embed fairness, transparency, and accountability into digital systems will not only comply with upcoming UK and EU standards, but also secure enduring public trust (DSIT, 2024; FCA, 2023).

---

References:

CDEI, 2021. The roadmap to an effective AI assurance ecosystem (Centre for Data Ethics and Innovation). [Online] Available at: https://assets.publishing.service.gov.uk/media/61b0746b8fa8f50379269eb3/The_roadmap_to_an_effective_AI_assurance_ecosystem.pdf [Accessed 15 December 2025].

DSIT, 2024. A pro-innovation approach to AI regulation: government response to consultation (Department for Science, Innovation and Technology). [Online] Available at: https://assets.publishing.service.gov.uk/media/65c1e399c43191000d1a45f4/a-pro-innovation-approach-to-ai-regulation-amended-governement-response-web-ready.pdf [Accessed 18 December 2025].

FCA, 2023. Sustainability Disclosure Requirements (SDR) and investment labels (Financial Conduct Authority). [Online]  Available at: https://www.fca.org.uk/publication/policy/ps23-16.pdf [Accessed 15 December 2025].

Floridi, L. et al., 2018. AI4People—An Ethical Framework for a Good AI Society: Opportunities, Risks, Principles, and Recommendations. [Online] Available at: https://doi.org/10.1007/s11023-018-9482-5 [Accessed 28 December 2025].

House of Lords, 2018. AI in the UK: ready, willing and able?. [Online] Available at: https://publications.parliament.uk/pa/ld201719/ldselect/ldai/100/100.pdf [Accessed 15 December 2025].

ICO, 2023. Guidance on AI and data protection (Information Commissioner's Office). [Online] Available at: https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/artificial-intelligence/guidance-on-ai-and-data-protection/ [Accessed 14 December 2025].

Leslie, D., 2019. Understanding artificial intelligence ethics and safety: A guide for the responsible design and implementation of AI systems in the public sector. [Online] Available at: http://doi.org/10.5281/zenodo.3240529 [Accessed 5 January 2026].

OECD, 2024. Recommendation of the Council on OECD Legal Instruments Artificial Intelligence (Organisation for Economic Co-operation and Development). [Online] Available at: https://legalinstruments.oecd.org/en/instruments/oecd-legal-0449 [Accessed 15 December 2025].

(Date: 9^th January, 2026)