Artificial intelligence has integrated itself into nearly every technology platform on the market, but Elon Musk’s Grok AI tool is just the latest litmis test to see how far tech companies will go with AI. Grok was recently embraced by the Pentagon for internal use after global outcry over users using the tool to “digitally undress” women and minors.
While its original intended purpose was educational and experimental, some users realized Grok’s potential for creating highly sexualized deepfake images, raising ethical and social concerns. In response to public criticism, Musk limited some of its image generation tools to paid subscribers.
“What made the Grok controversy significant is that it has never before been as easy to make an inappropriate picture of somebody,” computer science teacher Daniel Nawrocki said. “Since Grok is connected to Twitter, a lot of people saw it, and it made everyone question whether this kind of AI technology is good or bad.”
The controversy reveals a bigger issue: AI has gotten so realistic that the line between real and fake is blurring. What once seemed obviously computer-generated can now pass as authentic, making people second-guess everything they see online.
“One of the strengths is how realistic AI has become,” Nawrocki said. “You can take a picture of somebody and make them do something they never actually did, and people will believe it. That is also the weakness, because it can be used to make someone do something illegal or inappropriate without their permission. AI is only as good as the information it has been trained on. If it is trained on false or biased data, it will generate false or biased information. That is a major weakness of current AI systems.”
This growing realism has made it harder for people to trust online content. With every new viral post, the question isn’t just whether it’s true—it’s whether it’s even possible to tell.
“There is a huge lack of trust in the news now,” library assistant Tera Sheppard said. “People do not know what is real and what isn’t, so everything feels questionable. It really tests what people are willing to believe. Deepfakes are often created to hurt people, and they are very convincing. Because people already do not trust what they see, deepfakes push society into a place where no one knows how to decide what is real anymore.”
Even with careful attention, spotting AI-generated content can be tricky. Certain inconsistencies, like unnatural eye movements or extra fingers, can give away a fake—but not always.
“Usually when a video contains humans, I look closely at their eye movements because AI messes that up,” junior Jay Surappagari said. “It does not look very human. I also look at their fingers because AI tends to give people six fingers. Most humans do not have six fingers. That is usually how I can tell if a video is AI or not.”
Some videos are nearly flawless, though, making detection almost impossible. AI-generated baby videos, for instance, can be shockingly convincing.
“On Instagram, there are a lot of baby videos, and AI usually makes those videos really perfect,” Surappagari said. “The babies do not have six fingers, their eyebrows look normal, and they look like real babies. In those cases, it is very hard for me to identify if a video is AI or real.
In real life scenarios, deepfakes can cause real emotional and social harm, especially when they target individuals or sensitive events, like videos and images of ICE officers clashing with protesters. People are not just losing trust—they are experiencing personal impacts that can harm families, communities and public perception.
“For example, when Charlie Kirk was assassinated, there were a lot of deepfakes about him,” Surappagari said. “These deepfakes really hurt his family. When we look at the people who are making it, they are enjoying it and getting views. The reactions to these situations really depends on the audience and what type of content is deepfaked.”
The Grok controversy helps amplify the spread of deepfakes. By providing users with tools to generate highly realistic images quickly, it demonstrated just how rapidly AI technology can evolve.
“What surprised me most was how real deepfakes look now,” sophomore Bavishya Meruva said. “With the Grok controversy, paid users were able to create inappropriate images that did not even look AI-generated. Deepfakes have been evolving for years, but now they are so realistic that you honestly cannot tell the difference anymore.”
The rapid evolution of AI could have lasting impacts on young users. The accessibility of AI tools means that even children can generate material with long-lasting consequences, potentially affecting their futures in unexpected ways.
“It is 2026, and our world is ever changing,” Meruva said. “Things like what you put into ChatGPT between other AIs, it takes that feedback and puts it into different versions. It keeps evolving to be more and more human. If it is like this in 2026, imagine how it would be in the future just a year from now. That is unimaginable. It is not just accessible to adults, because it is really not hard to use. You just need to put in a prompt and you can get anything from it. That means even kids can use it. It is also very permanent because if a picture was leaked of you, your entire future could be ruined, even if you are just 13.”
As AI tools expand across social media, the Grok scandal underscores the growing gap between innovation and oversight, raising questions about accountability, privacy and the future of online trust.