Trust
AI governance.
AI is part of how GreenBDG Africa works. It validates meter data, supports analysis and speeds up reporting across our operations, products and services. This policy sets out the principles, responsibilities and controls that govern how we develop, deploy and use AI, so that it is used ethically, responsibly, transparently and in compliance with applicable law and industry best practice.
Four commitments matter most to the people we work for:
Human review before release
No AI-generated recommendation, report or disclosure reaches you unreviewed. Qualified people check every material output before it goes out.
Your data stays protected
Confidential client information does not go into public AI tools. Every AI application must comply with POPIA and our data classification controls.
You will know when AI contributed
Clients are told when AI materially contributes to analysis, reporting or recommendations, and outputs are traceable to their sources wherever possible.
AI's own footprint is managed
We use AI efficiently, minimise unnecessary computational demand and prefer providers with credible renewable energy and climate commitments.
The policy in full
01 Purpose
This policy exists to:
- accelerate sustainability outcomes and improve operational efficiency through responsible AI adoption;
- ensure AI is used ethically, transparently and in compliance with applicable laws and industry best practice;
- maintain human oversight so AI supports, rather than replaces, professional judgement;
- protect data privacy, confidentiality and the integrity of ESG reporting and client advice;
- manage AI-related risks (privacy, cybersecurity, regulatory, reputational, bias and environmental) in a structured way;
- minimise the environmental footprint of AI use and favour providers with strong sustainability credentials.
02 Scope
This policy applies to:
- People. All GreenBDG Africa employees, contractors and consultants.
- Internal platforms. AI-enabled products and services, including AfricaESG.AI, SUMS and associated monitoring, reporting and analytics platforms.
- Third-party tools. Generative AI applications used in business operations, such as ChatGPT, Microsoft Copilot, Gemini and similar technologies.
- Systems. AI systems developed internally or procured from external providers.
03 AI governance principles
GreenBDG Africa commits to the following principles:
3.1 Human oversight
- AI supports human decision-making and does not replace professional judgement.
- All material recommendations, reports, ESG disclosures, sustainability assessments and compliance-related outputs generated using AI must be reviewed by qualified personnel before release.
3.2 Transparency
- Users and clients are informed when AI materially contributes to analysis, reporting or recommendations.
- AI-generated outputs must be traceable to their sources and assumptions wherever possible.
3.3 Accuracy and reliability
- AI outputs must be validated before they are used for client advice, compliance reporting or investment decisions.
- Employees may not rely solely on AI-generated information without independent verification.
3.4 Fairness and non-discrimination
- AI systems shall not intentionally create or reinforce unfair bias based on race, gender, disability, nationality or other protected characteristics.
- Where AI models support decision-making, we periodically assess potential bias and fairness impacts.
3.5 Privacy and data protection
- AI applications must comply with applicable data protection legislation, including the Protection of Personal Information Act (POPIA).
- Employees may not upload confidential client information, personally identifiable information, financial data or proprietary project information into public AI platforms unless approved controls are in place.
3.6 Sustainability
- We use AI efficiently and minimise unnecessary computational demand.
- We prefer providers with strong renewable energy and climate commitments.
- We measure and disclose material AI-related emissions where feasible.
04 Acceptable use of AI
Employees may use approved AI tools for:
- research support;
- drafting reports and presentations;
- data analysis;
- coding assistance;
- sustainability modelling;
- ESG reporting support;
- operational efficiency improvements.
All AI-generated content must be reviewed for accuracy, completeness, bias, compliance and confidentiality.
05 Prohibited uses
Employees shall not:
- upload confidential client information to unauthorised AI platforms;
- use AI to generate fraudulent, misleading or deceptive content;
- allow AI systems to make unsupervised decisions affecting clients;
- present AI-generated outputs as verified facts without validation;
- use AI in ways that violate intellectual property rights or contractual obligations.
06 Data governance
All AI applications must comply with GreenBDG Africa's data management standards. Data used for AI purposes is classified by sensitivity:
| Classification | Permitted AI use |
|---|---|
| Public data | May be used within approved AI platforms. |
| Internal data | May only be used in approved enterprise AI environments. |
| Confidential client data | Requires explicit authorisation and approved security controls. |
| Restricted data | May not be entered into any public AI system. |
07 AI risk management
AI-related risks are assessed across seven categories:
- privacy risks;
- cybersecurity risks;
- ESG reporting risks;
- regulatory risks;
- reputational risks;
- bias and fairness risks;
- environmental impacts.
High-risk AI applications must undergo formal review prior to deployment.
08 Governance structure
| Role | Responsibility |
|---|---|
| Executive Committee | Maintains overall accountability for AI governance. |
| AI Governance Lead | Appointed by the CEO. Responsible for policy implementation, risk management, compliance monitoring, training and awareness, and vendor assessment. |
| Business units and all employees | Remain accountable for responsible AI use within their respective functions. |
09 Third-party AI providers
GreenBDG Africa assesses AI vendors against six criteria:
- data privacy compliance;
- security controls;
- transparency standards;
- ESG commitments;
- regulatory compliance;
- service reliability.
Vendor contracts include appropriate safeguards regarding data handling and intellectual property.
10 Training and awareness
All employees receive periodic training on:
- responsible AI use;
- data privacy obligations;
- AI risks and limitations;
- prompt engineering best practices;
- verification of AI-generated outputs.
11 Monitoring and review
This policy is effective from 19 June 2026 and is reviewed annually, or whenever significant technological, legal or business developments occur. GreenBDG Africa continuously monitors emerging AI regulations and international standards, including:
- ISO 42001, Artificial intelligence management systems;
- the OECD AI Principles;
- the UNESCO AI Ethics Recommendations;
- South African data protection requirements;
- evolving ESG disclosure frameworks relating to AI.
12 Policy statement
AI should be deployed as a force multiplier for sustainability, resilience and responsible development. Through sound governance, transparency and human oversight, GreenBDG Africa seeks to harness AI in a manner that creates lasting value for clients, communities and the environment, while managing associated risks responsibly.
Questions about our AI governance: info@greenbdgafrica.com.