How We Use AI Tools
Carefully, Usefully, and With Your Data Protected
How We Use AI Tools - Carefully, Usefully, and With Your Data Protected
We use AI tools to help us work faster and think more clearly — for example, to draft first versions of content, research general topics, troubleshoot technical questions, and support our Salesforce work. They help with the groundwork, but they do not replace our judgement, experience, or responsibility.
We know many nonprofits handle sensitive information about supporters, service users, staff, volunteers, funding, and internal plans. That is why our approach is simple: we only use AI where it is helpful, we stay in control of what information goes into it, and we keep client data inside protected tools covered by proper business terms.
What we use AI for
- Research and first drafts — we may use AI to help with background research, plain-English explanations, troubleshooting, and early drafts where client data is not needed.
- Salesforce delivery work — we may use AI to help draft formulas, flows, validation rules, documentation, and similar configuration work, but everything is still reviewed, tested, and owned by a human consultant before it is delivered.
- Understanding Salesforce orgs — we use a specialist Salesforce analysis tool to help us understand setup, such as fields, flows, validation rules, and automation logic. It helps us review configuration more efficiently.
- Salesforce AI features — Salesforce has its own AI features, such as Agentforce, which sit within your Salesforce environment and are governed by your own Salesforce terms and settings. If you ask us to help enable or configure them, we will talk through the practical implications with you first — we would never simply switch them on without discussion.
How we protect your data
We treat confidentiality as a practical responsibility, not just a clause in a contract. If information belongs to you, your organisation, your supporters, or the people you work with, we handle it with care.
In practice, this means:
- Client information stays inside protected business tools. Where AI is used with information from emails, documents, or project work, we use Microsoft 365 Copilot in our business environment, where normal Microsoft 365 permissions, security controls, and data protection terms apply.
- We do not put client data into free or consumer AI tools. Those tools can be useful for generic tasks, but we do not use them for anything that includes your confidential information, personal data, internal documents, Salesforce records, or client-specific details.
- Salesforce org analysis is limited to setup, not records. When we use a specialist Salesforce analysis tool, it is used to understand how your system is configured — not to analyse the personal data held in your records. The tool we currently use is ISO 27001 certified, which means its information security management system is independently audited against a recognised international standard. We can share further information about the tool and its security with clients on request.
Working with trusted people
Sometimes we work with trusted contractors. When we do, they are covered by the same confidentiality expectations as our own team. If they use AI tools in connection with client work, those tools must be appropriate business products with clear protections around confidentiality, data use, and model training — including safeguards so client information is not used to train public or general-purpose AI models.
Our commitment
AI can be genuinely useful, but it still needs human care around it. We check AI-assisted work, test what we deliver, and use our own judgement before anything reaches you or your users.
This policy may be updated from time to time as AI tools and best practice evolve. The version on this page is always the current one. If you ever want to understand how AI has supported a particular piece of work, just ask — we are always happy to explain our approach.
Last Updated 9th July 2026