How Not to Use AI: Stanford University Can Tell You

Stanford had an AI rule. Then it broke its own AI rule.

The photograph looked like an ordinary college promotional image: three Stanford students smiling and holding plates of food at a Lunar New Year celebration.

It wasn’t an AI-generated image. It was a real photograph of three real students.

Then Stanford changed it.

What Happened to the Photograph?

The original photograph was taken at Stanford’s Wilbur Dining Hall during a 2024 Lunar New Year celebration, when Billy Ramirez was a freshman.

The photograph had previously been used on Stanford’s residential and dining website.

Later, Stanford Residential & Dining Enterprises (R&DE), the department responsible for Stanford’s housing and dining services, used AI to modify the photograph for a promotional banner.

Ramirez, now a senior, was removed from the photograph entirely and replaced with an AI-generated Black woman wearing Stanford apparel. The change altered the person’s race, gender and appearance.

The other two students remained in the photograph, but their appearances were changed as well. Their faces were altered, and they appeared thinner in the altered image. Their clothing was changed to Stanford apparel.

A friend sent Ramirez a side-by-side comparison while he was driving back to campus.

At first, he laughed.

Then he started thinking about what had actually been done to the photograph.

He told the San Francisco Standard:

“They wanted to show the world, ‘Oh, this is what Stanford looks like,’ when it’s really just all fabricated.”

A real student had been removed from a photograph and replaced with an AI-generated person who does not actually exist.

The other students in the photograph had been changed, too.

And nobody looking at the promotional image was told that it had been altered with AI.

Stanford’s Own Policy Prohibited It

Stanford’s AI policy prohibits using AI to alter images of Stanford people, events, research, facilities, or achievements.

The university also has guidance calling for disclosure of AI-generated or AI-manipulated content when failing to disclose it could mislead people.

In this case, Stanford confirmed that both the alteration and the failure to disclose it violated its policies.

A university spokesperson told The Stanford Daily that the AI alteration violated Stanford’s policy.

Stanford removed the banners and said it would provide additional staff training and ensure proper review of materials.

What Does This Mean for Your Website?

The same issue can come up on any website.

Maybe you use AI to create a promotional image for your business. Maybe you modify a photograph of your employees, your location, your products or an event.

There’s a place for AI-generated illustrations or graphics that aren’t meant to represent real people or events.

You might also use AI to make a simple adjustment to a real photograph, such as changing brightness or contrast.

The situation changes when you take a real photograph and substantially change it.

If you change a person’s clothing, alter their appearance, add people who weren’t there or replace someone with an AI-generated person, you’re no longer just cleaning up the photograph.

You’ve changed what the photograph shows.

If visitors would reasonably assume they’re looking at an authentic photograph, it’s worth telling them that it was altered.

It doesn’t have to be complicated.

“Image digitally altered using AI.”

That may be all the explanation you need.

The Person Publishing It Still Has to Look at It

AI can make the changes in seconds. That doesn’t mean you should automatically publish whatever comes out of the tool.

Someone still has to look at the finished image and decide whether it accurately represents what you intend it to represent.

Is it a fictional image? Fine.

Is it an edited photograph of something real? Then look at what was changed.

If the changes are substantial enough that someone could mistake the image for an authentic photograph, disclose the alteration.

Stanford’s situation is a particularly clear example because its own policy addressed both parts of the problem: the alteration itself and the failure to disclose it.

The technology made the image possible. A person still decided to use it.

That’s the part AI can’t decide for you.


Not a fan of AI? Read: Looking for a Website Builder Without AI? Which One Should You Choose?

UltimateWB uses software – not AI – to help you design and build your own website. Beginner friendly, expert flexible – Learn more about UltimateWB! We also offer web design packages if you would like your website designed and built for you.

Got a techy/website question? Whether it’s about UltimateWB or another website builder, web hosting, or other aspects of websites, just send in your question in the “Ask David!” form. We will email you when the answer is posted on the UltimateWB “Ask David!” section.

About the UltimateWB Team

This article was written and reviewed by the UltimateWB Development Team. With over 20 years of hands-on experience in full-stack web development, database optimization, and secure server administration (WHM/cPanel), we engineer UltimateWB with clean, built-in apps so you never have to deal with the performance-draining software bloat, security risks, or compatibility issues of third-party plugins. We build software designed from day one for maximum developer autonomy and lightning-fast performance.

This entry was posted in AI, Technology in the News and tagged , , , , , , , , , , . Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *