Responsible AI and Data Use statement
AI for Non-profits
Last updated: 14 September 2026
Our approach
We use AI where it supports useful, proportionate work and is consistent with our values: care, dignity, inclusion, accountability and community trust.
We assess AI use by its purpose, the information involved and its potential impact on people. AI supports our work; responsibility for our decisions and the quality of our services remains with us.
How we use AI
We may use AI to support:
- drafting, editing and structuring content
- summarising public or low-risk information
- organising ideas, notes and actions
- developing learning materials and practical resources
- reducing repetitive administrative work.
We choose tools according to the task and level of risk. We do not use AI where a simpler approach would be more appropriate.
We do not use AI to make safeguarding decisions, replace professional judgement or produce unchecked advice or recommendations.
Confidentiality and data protection
We treat confidentiality as a responsibility to the people and organisations we work with.
Safeguarding information, case-related material and other highly sensitive information must not be entered into third-party AI tools.
For other work, we limit information to what is necessary and remove identifying or confidential details before using AI. Removing names alone may not be sufficient: contextual details can still identify a person or organisation.
Where information cannot be used safely, we do not upload it. Our use of AI must respect applicable data protection requirements, confidentiality commitments and any restrictions agreed with clients or partners.
Human review and accountability
AI-generated content can contain errors, unsupported claims, bias and misleading interpretations. We do not treat confident language or polished presentation as evidence of accuracy.
Before AI-assisted work is published, shared externally or used to inform a significant decision, we review it for:
- factual accuracy and support from original sources
- relevant context, uncertainty and differing perspectives
- confidentiality and privacy risks
- bias, unfairness and potential harm
- clarity, accessibility and suitability for its audience.
We remain responsible for the work we deliver, including work supported by AI.
Transparency
We are open about meaningful uses of AI in our work. Where AI materially contributes to a service or output, we explain its role and discuss relevant boundaries with the client or partner.
Routine uses, such as formatting or low-risk internal brainstorming, do not require individual disclosure. We welcome questions about our approach and respect agreed restrictions on AI use.
Tool selection and access
Before using an AI tool, we consider its purpose, suitability and data-handling arrangements, including how information may be stored, retained, shared or used for training.
We also consider the permissions it requires and whether it connects to files, email or other systems. Access should be limited to what is necessary for the agreed task.
We do not use tools with confidential material where we lack sufficient assurance about how that material will be handled.
Fairness, dignity and inclusion
AI systems can reproduce prejudice, misrepresent people and overlook experiences that are poorly represented in their training data.
We review outputs for discriminatory assumptions and exclusionary language, including those relating to race, disability, gender, sexuality and other aspects of identity or experience.
We preserve people’s own accounts and distinguish evidence from interpretation. We do not use AI to fabricate views, testimonials or lived experience, or to present generated content as authentic consultation or participation.
Environmental responsibility
We recognise that AI has environmental costs. We aim to use it proportionately, avoid unnecessary generation and choose less resource-intensive approaches where they meet the need.
We do not make environmental claims that we cannot substantiate.
Review and improvement
We review this policy as our work, tools and understanding of AI develop. Where we identify an error or concern in AI-assisted work, we assess its impact and take appropriate corrective action.
Our guiding commitments remain consistent: protect people, respect confidentiality, apply human judgement and take responsibility for our work.
Questions about this policy or our use of AI can be raised through the contact details on our website.