Artificial intelligence (AI) is developing rapidly, and charities are increasingly – and understandably – exploring how it can support their work. From helping to draft fundraising communications and analyse data, to creating images and providing automated responses to donors, AI has the potential to bring significant benefits.
But as its use grows, so does the need to think carefully about how it is used. For fundraisers, the question should not simply be whether AI can make something quicker, cheaper or more effective. It should also be whether its use is consistent with the Code of Fundraising Practice and the standards the public expects from charities.
Public trust is hard earned and easily lost. That means responsible use of AI should start with the same guiding principles that underpin the code: legal, open, honest and respectful.
Our guidance on using AI in fundraising is designed to help charitable fundraising organisations apply the code when exploring, preparing for and using AI. It recognises that there is no one-size-fits-all approach. The level of oversight and transparency needed will depend on what AI is being used for and the risks involved.
Assess the risks, test, and innovate responsibly
AI presents many opportunities for charities. A fundraiser might, for example, use AI to help produce a first draft of a campaign email, generate ideas for a fundraising event or analyse patterns in donor behaviour. Used appropriately, these tools can support fundraisers in building engagement and generating donations.
But AI can also produce information that looks convincing but is wrong or harmful. It can reproduce hidden biases, generate inappropriate content or produce outputs that unintentionally misrepresent people or situations. There are also important questions around personal data, intellectual property, cybersecurity and the use of AI-generated images, audio and video.
Consider these two examples.
A charity uses AI to draft the content of a fundraising appeal and suggest how best to communicate it through different communication channels. A member of staff checks every fact, carefully considers the wording and tone, and approves the final version. The charity has assessed the tool’s risks and understands how it handles data, and publishes an AI policy. In this scenario, AI supports human judgement, it does not replace it.
Now consider the opposite: a charity uses AI to identify potential donors and automatically target fundraising materials, without knowing how the system makes decisions or checking whether they are accurate or fair. It may also upload personal information to a public AI tool without knowing how that data will be used or stored. This approach puts donors and public trust at risk.
Our guidance is clear: Before using AI, organisations should carry out a proportionate risk assessment, develop an AI policy (published on your website), ensure staff have the necessary skills, and test or pilot tools before introducing them more widely.
AI suggests, people decide
Strong oversight and safeguards are essential. AI should not become a way of transferring responsibility from people to technology. Fundraising organisations remain accountable for the content they produce and the decisions they make, including where AI has contributed to them.
This is particularly important when organisations work with third parties. A charity cannot assume that responsibility for meeting the code sits elsewhere simply because an AI tool is provided or operated by a supplier or fundraising partner. As with all third-party relationships, appropriate checks and ongoing monitoring remain vital.
Ask the right questions
Before introducing AI, fundraisers should consider the following questions, alongside any other considerations relevant to their organisation: What are we using it for? What could go wrong?
What information will we put into it? Could its outputs mislead, discriminate against or disadvantage someone? Who will check the results? What will we tell donors about its use? And how will we know whether the tool continues to work as intended?
Recording decisions and the checks undertaken can also help organisations demonstrate that they have considered the risks and acted appropriately if concerns are later raised.
The code’s principles-based approach is designed to support this kind of judgement. It gives fundraisers flexibility to respond to new and emerging methods and technologies even when they are not specifically mentioned. Our AI guidance should be used alongside the code to help organisations make those judgements.
A shared responsibility
Charities are not the only organisations shaping how AI is used in fundraising. Organisations that advise, support, train or provide information to the charity sector also have an important role to play.
We encourage all organisations promoting or advising on the use of AI in the charity sector to acknowledge our guidance and, where appropriate, integrate its principles into their own resources, training and advice. This will help ensure that fundraisers receive consistent messages about the importance of accountability, transparency, risk assessment and human oversight.
AI will continue to evolve, and fundraising will evolve with it. The sector should be open to the opportunities this creates. But innovation should never come at the expense of the public trust and confidence that makes charitable fundraising possible.
Paul Winyard is director of policy at the Fundraising Regulator
