This article was published based on recent guidance written by Kieron James (Board Member) regarding guidelines for using artificial intelligence in charity fundraising. This guidance comes at a moment when charities are rapidly adopting AI tools, making the real question not: should we use the technology or not? But rather: how do we use it without touching the core of trust upon which charity work is based? The article is based on key guiding ideas: the “lifecycle” approach to adopting AI, the board's responsibility for decision-making and accountability, the importance of training, the principle of proportional transparency depending on the risk of misinformation, and the necessity of human oversight to ensure accuracy, fairness, and compliance. The goal here is to transform these guidelines from a regulatory document into a practical read that helps professionals in the nonprofit sector to employ AI as a tool to enhance efficiency without compromising integrity, respect, and clarity.

As Tools Accelerate, Difficult Questions Slow Down

Artificial intelligence (AI) is no longer just an “additional tool” in the fundraising toolbox; it has become part of the daily landscape in writing messages, content creation, data analysis, improving targeting, and perhaps even designing more personalized communication experiences with donors. This rapid expansion opens an important door to enhance efficiency and reduce effort during times of financial and operational pressure, but at the same time, it raises more sensitive questions: how do we ensure that AI tools do not inadvertently mislead donors? How do we prevent “persuasive” but inaccurate content? What is the role of the board and executive leadership in steering the direction before technology turns into a reputational and trust risk?

In this context, the issuance of new guidance explains how to use AI in fundraising in accordance with the approved Code of Fundraising Practice. The core idea is not to encourage or discourage the use of AI, but to set a clear standard: if it is used, it must be legal, open, honest, respectful, and compliant with professional standards.

Why Now? High Adoption Rates and Urgent Need for Regulation

Data from 2025 indicates that the percentage of charities using artificial intelligence has increased compared to the previous year, alongside a growing sense in the sector of the need for practical help in understanding “how to comply” when introducing these tools into fundraising. This finding reveals a crucial point: technology adoption often precedes the accompanying organizational maturity. A charity may start using text generation or data analysis tools before setting its internal policies or before the board understands the nature of the risks and limits of responsibility.

The new guidance steps in to fill a practical gap: not by providing technical rhetoric, but by providing a professional oversight roadmap that views AI as a “process” with stages and responsibilities, rather than as a program we load and then embark.

The “Lifecycle” Approach: AI is Not a Snap Decision

One of the strongest points of the guidance is its adoption of a “lifecycle approach” to dealing with artificial intelligence. This means that it does not only target those already using the tools but also includes three scenarios:

  1. Exploration Stage: Can AI support fundraising? What are the possible uses? What are the risks?
  2. Readiness Stage: Choosing the tool, designing the process, identifying who reviews, and setting transparency rules.
  3. Operational Use Stage: Running the tool within campaigns, monitoring its impact, and managing deviations and errors.

Importantly, the approach is not limited to the charity’s direct use of the tool; it also includes dealing with third parties (vendors, agencies, platforms) that may use AI on their behalf. A charity may not plan to use AI, but a contractor may use it to write messages, design content, or analyze targeting data, which means that “risks may enter from outside” even if the internal policy is conservative.

The Role of the Board: Overall Responsibility Does Not Shift to the Tool

The guidance emphasizes that trustees or the board are the top accountable party for the use of AI in fundraising. This changes the perspective: the matter is not merely a technical decision made solely by executive management; it is a strategic decision that affects trust, compliance, and reputation.

However, the board's accountability is not formal. What is required is “empowering understanding”; that is, the board must build sufficient knowledge to realistically evaluate opportunities and risks. This requires specific governance questions, such as:

  • What are the limits of using AI within the charity? Where do we prohibit it?
  • What level of risk is acceptable? Who decides?
  • What is the mechanism for human review before publication or sending?
  • Do we have a clear policy for dealing with errors? How do we report if unintended misinformation occurs?
  • What about suppliers? Do we have a contractual clause requiring transparency and oversight?

The crux here is that the board does not need to become a “technical team,” but it needs to understand AI well enough to protect the charity from short-sighted operational decisions.

User Training: Skill is an Ethical Requirement Before Being a Technical One

The guidance clearly indicates that training is a central element: anyone who develops, manages, or uses AI tools in fundraising must have sufficient skills. In the nonprofit sector, “quick-fix” temptations may arise: an employee copies text from a tool, requests a compelling story, or generates a mailer in minutes. However, a lack of skill may lead the charity to fall into three common gaps:

  • Accuracy Gap: Text that seems excellent but contains incorrect information or undocumented numbers.
  • Relevance Gap: Language that does not fit the charity's values or its local context or the sensitivity of its audience.
  • Compliance Gap: Content that may exaggerate claims or imply misleading information to the donor unintentionally.

Training here is not a “technical course,” but building the ability to distinguish between publishable content and content that needs rewriting, documentation, and review.

Transparency: Not a Permanent Sign… But a Decision Proportional to the Risk of Misinformation

One of the most sensitive points is transparency when using artificial intelligence. The guidance establishes a balanced principle: not all uses of AI need a label or disclosure. Sometimes this is unnecessary or disproportionate to the nature of the use. However, in other cases, concealment may mislead the donor or the public.

Examples of highly sensitive cases include:

The proposed standard is clear: the greater the likelihood of misinformation, the greater the need for transparency.

  • AI-generated content that mimics a personal testimony or human experience that may be understood as real.
  • Images, stories, or “quotes” that may mislead the recipient about a non-existent reality.
  • Messages suggesting that some communication was made based on accurate knowledge of the donor's circumstances or intentions, when in fact, it is the result of machine analysis that may be wrong.

Transparency here is not a luxury. It is a tool for protecting trust before being merely a formal obligation.

Proportional Human Oversight: Because AI May Err Confidently

The guidance warns that AI tools may generate “convincing yet incorrect” content, reproduce biases, or cause unintended consequences. Therefore, it emphasizes the necessity for proportional human oversight to review: accuracy, fairness, and legal compliance before using any content in fundraising.

“Proportional” here means that the level of scrutiny increases as the sensitivity of the use increases. A short general message may require linguistic editing and basic fact-checking, while a major campaign or beneficiary story or impact figure announcement may need multi-level reviews, possibly source documentation, leadership approval, and compliance checks before publication.

Here, AI becomes an assistant, not a substitute: it speeds up drafts, but does not determine truth.

The Ethical, Social, and Environmental Dimension: Does the Use Align with the Charity's Values?

The guidance also points out that some charities are thinking about a deeper issue than efficiency: does the use of artificial intelligence align with our values? There are ethical, social, and environmental considerations associated with this technology. The decision may not only be “permissible” but also “appropriate” to the charity’s identity and mission. This is a mature question that should be raised within boards, as the value of nonprofit work is not measured solely by outputs but by trust, meaning, and ethical behavior.

AI Does Not Threaten Charities… Misuse Is the Threat

The new guidance delivers a realistic message to the sector: you do not have to be either a proponent or an opponent of AI. What matters is managing its use according to a professional standard that protects the donor and safeguards the charity. The lifecycle approach reminds us that technology is an ongoing decision, and the role of the board ensures that governance precedes enthusiasm, proportional transparency prevents confusion, and human oversight limits the “hallucination” errors and biases before they turn into a reputational crisis.

In the end, artificial intelligence can enhance fundraising efficiency, but it will not buy trust. Trust is built when the charity is accurate in its story, honest in its messaging, clear in its intentions, and governed by standards that put technology at the service of the message, not in leading the message.