BALTIMORE — At first glance, the videos looked like tributes.
Older men wearing veteran-style hats, American flags, emotional captions, voices asking viewers to stop scrolling, say thank you or offer support.
But after WMAR-2 News reviewed several Facebook pages sharing veteran-related images and reels, questions emerged about whether some of the content showed real veterans and real stories — or whether it was created or assisted by artificial intelligence.
WMAR-2 News reached out to Meta, Facebook’s parent company, about three pages sharing veteran-related content. Meta said the pages have since been removed.
“We removed these pages for violating our policies,” Meta said in background information provided to WMAR-2 News.
The pages WMAR-2 News asked about included:
- https://www.facebook.com/veteransstoriesofficial
- https://www.facebook.com/strong0america/reels/
- https://www.facebook.com/profile.php?id=61588860848138&sk=reels_tab
Meta did not directly confirm whether all of the content on the pages was AI-generated. However, the company said labeling AI-generated content across its platforms can be difficult.
“It’s not always possible to label every AI-generated image, video, or audio content shared on our platforms because many companies haven’t added industry-signals to video and audio generators. Additionally, some AI content doesn’t include signals, and people find ways to remove or hide signals that would otherwise be used to apply AI labels to posts. Identifying AI-generated content can be challenging as the technology evolves. We’re continuously working to improve our systems and ability to label this content. That’s why it’s important to consider several factors when determining if content has been created by AI. This includes checking the trustworthiness of the account sharing the content and/or looking for details that might look or sound unnatural,” Meta said.
Daniel Trielli, an assistant professor of media and democracy at the University of Maryland Philip Merrill College of Journalism, reviewed examples of the veteran-related content WMAR-2 News had been tracking.
“These videos are often tailor-made for us to have emotional reactions,” Trielli said.
Trielli said some common warning signs include glossy or overly polished visuals, slightly unnatural movement, repeated video formats and captions designed to make people like, share, comment or follow.
He said veteran-related content can be especially powerful because it taps into service, patriotism and a shared sense of community.
“People who don’t know each other but discover that they’re veterans suddenly have some sort of a bond that creates that community,” Trielli said.
Trielli said the issue is not that all artificial intelligence is bad. AI tools can be used for captions, accessibility, research, translation and storytelling. The concern, he said, is when realistic-looking images, voices or stories are created or edited with AI and viewers are not told what they are seeing.
“AI is a big field and it can have good uses,” Trielli said. “But as every technology, it can have very bad uses too.”
When AI-generated or fictionalized content is not clearly labeled, Trielli said it can create several risks. A post could mislead someone into donating money, sharing personal information or following a page that later promotes scams or misinformation.
But he said the goal is not always immediate money. Sometimes, it is attention.
“If you follow a page, if you like a page, the algorithm keeps pushing that page,” Trielli said.
Trielli said even skeptical comments can help boost a post because platforms often treat comments, likes, shares and watch time as engagement.
That means arguing in the comments may still help the post spread.
For viewers, Trielli said the first step is to slow down before sharing.
Look for basic details: a real name, date, location, military unit, original source or organization behind the post. Be cautious if a page shares emotional military content but does not identify who is pictured, where the story came from or whether it is real, recreated or AI-generated.
“The best thing that you can do is fight that urge, really look at that image and think about it,” Trielli said.
He also said AI detection tools can help, but they are not perfect. They can miss AI-generated content or falsely label real content as AI.
“They can either give you false negatives… or give you false positives,” Trielli said. “I wouldn’t rely on them 100%.”
Instead, Trielli said viewers should use critical thinking, check the source and report suspicious content to the platform.
“You as a user have a lot of power,” Trielli said. “You can report that content, but you could also just close it.”
Meta says users can report content they believe violates Facebook policies.
The takeaway, experts say, is not that every use of AI is harmful. It is that transparency matters — especially when content involves real communities, military service and personal sacrifice.
This story was written with the assistance of artificial intelligence and reviewed by WMAR-2 News staff.
