AI image generators can produce campaign graphics, social media images and website illustrations within seconds. For nonprofits operating with small communications teams and limited design budgets, that capability is attractive.
However, using AI-generated images is not merely a design decision. It can create ethical problems involving consent, representation, accuracy and public trust.
Nonprofits should establish clear boundaries before synthetic images become a routine part of their communications.
The Risk of Inventing the People You Serve
Consider a nonprofit that provides housing assistance. Instead of photographing an actual program participant, the communications team generates an image of a distressed family standing outside an apartment building.
No real person was exploited or identified. On the surface, that may appear more ethical than using a beneficiary’s photograph.
The problem is that the organization has created fictional evidence of its work.
Supporters may reasonably assume that the image shows actual people, an actual location or an actual outcome. Even when the underlying message is accurate, the visual presentation can become misleading.
Synthetic images are especially risky when they depict vulnerable populations, including children, refugees, people experiencing homelessness, domestic violence survivors or people receiving medical care.
The central question should be simple:
Could a reasonable viewer mistake this image for documentation of a real person, event or program?
When the answer is yes, the image should be clearly labeled, redesigned as an obvious illustration or not used.
Consent Still Matters
AI-generated images do not eliminate consent concerns.
Some tools can recreate a person’s face, voice, body or artistic style. Others may produce images that closely resemble real individuals without their knowledge.
The UNESCO Recommendation on the Ethics of Artificial Intelligence emphasizes human dignity, privacy, fairness, transparency and accountability. These principles apply directly to nonprofit communications.
An organization should never use AI to alter a beneficiary’s appearance, manufacture a testimonial or place someone into a fictional situation without explicit permission.
Consent should also be specific. Permission to use a photograph in an annual report does not automatically grant permission to alter that photograph with AI or reuse the person’s likeness in a generated campaign.
Representation and Bias
Image generators learn patterns from enormous datasets. Those patterns can reproduce cultural stereotypes.
A prompt requesting an image of poverty may consistently generate certain racial groups. A request for nonprofit executives may produce mostly white men. Images of people with disabilities may appear medically inaccurate or patronizing.
Communications teams must review more than visual quality. They must examine what the image suggests about dignity, competence, race, gender, age, disability and economic status.
The goal is not simply diverse imagery. The goal is responsible representation.
The human-rights organization WITNESS offers guidance for responsible synthetic media, emphasizing consent, disclosure and protection for communities that may be harmed by manipulated content.
Create a Clear Approval Process
Nonprofits should document when AI-generated images are acceptable.
A practical policy might allow synthetic images for abstract concepts, decorative backgrounds, internal presentations and clearly illustrated educational material.
It might prohibit them for:
- Beneficiary stories
- Program documentation
- News coverage
- Testimonials
- Emergency appeals
- Impact reporting
- Images resembling identifiable people
Every public image should have a staff owner responsible for reviewing its accuracy, permissions and context.
This is where an AI Knowledge Hub such as Maisy becomes useful. The organization’s image policy, consent requirements, brand rules and approval procedures can be stored in one governed location. Employees can ask questions such as, “Can we use an AI-generated child in this fundraising campaign?” and receive an answer based on approved organizational policy rather than personal assumptions.
AI-generated images can reduce costs and expand creative options. They can also damage trust quickly.
For nonprofits, the correct standard is not whether an image looks convincing. It is whether the image is honest, respectful and consistent with the mission.
