Should You Watermark AI-Images You Create?

AI images with a sample watermark to illustrate the concept on nonprofit's watermarking images

Should nonprofits disclose when they incorporate AI (artificial intelligence) into their content creation for communications, marketing, and fundraising? When is AI a simple tool you don’t disclose, like using spelling and grammar checkers? When is it an original content creator whose “authorship” should be documented?

As of today (in May 2024), we at Nonprofit Marketing Guide believe nonprofits should disclose when they use AI to create photographic images by placing a custom watermark on the image.

Photographs are powerful ways to communicate about your organization’s needs and solutions. But nonprofits often don’t have direct access to people and places they might otherwise want to photograph. Protecting the privacy of the people you serve often prevents using real photography.

But no one wants to be fooled into thinking they are looking at something real when it’s actually been created by a machine.

Creating a Custom Watermark for AI Images

So, how can you use the power of AI-generated photography while still being transparent with your supporters? One approach is to apply a watermark to your content.

Many AI companies, search engines, stock photo companies, and others interested in image provenance are already working on ways to add digital watermarks to images that can’t be seen by the human eye. This is especially important given the malicious use of deepfakes.

But what we suggest nonprofits add to their AI-images would be visible to the human eye. Much like you might already overlay your logo onto an image, you would create and then overlay a disclosure watermark.

What might that watermark for your nonprofit organization look like? We asked participants in our recent webinar on AI policy for nonprofits for their ideas. Here are a few concepts to get you thinking: