Photos and videos are getting too hard to differentiate what’s real and fake

With recent Artificial Intelligence (AI) tools, such as DALL-E image generation, a tool created by the AI research lab, “OpenAI” and Chat-GPT 5.2. According to Mashable, it was released on Dec. 11, 2025 with even more advanced image generation. In doing so, we can possibly be seeing an even larger gap between differentiating real photographs compared to images created by AI computers. It’s safe to say that soon, creative hobbies such as being an independent artist might become obsolete, as ChatGPT may improve further, taking out all the time and enjoyment of creation, such as using digital platforms or pen and paper and changing it to typing in a sentence and then having no attachment to the creation at all and now it is making photos that are indistinguishable from those created by a human. As we’ve been seeing, AI is pouring into our photos and it was made clear that it will make its debut in films as well. For example, the program Sora 2 AI released on Sept. 30, 2025 creating incredibly advanced videos from a singular prompt. However, these tools are not unlimited as Sora 2 can only generate 15-second videos,  and if you want to generate 25 second videos, you need to shell out $200 a month.

With these tools getting more advanced however, it’s becoming more difficult to differentiate what’s real and what’s fake, although there are a few silver linings, such as its inability to generate human hands correctly, commonly giving hands six or four fingers. Another thing that it has trouble with is creating text, shadows are very complex to make as well which is another way to tell a fake photo from a real one.

Generative AI tools are somewhat divisive, Bear Ogletree ‘29 said, “They’re probably some of the worst things ever made. The technology shouldn’t be advanced any further as well. Sometimes I might use it if I need help with writing or whatever, but not really.”

 Idriss Snoussi ‘27 said, “I think that they [AI generated photos and videos] are funny and fine, until they start to take jobs and start causing drama, besides that I think it’s fine. I think that the technology will advance, and I think that it should advance to the point where it can detect if you’re trying to generate a specific person or like a politician.”

For what we are seeing and receiving, the AI generative technology is advancing quickly, as DALL-E began Jan. 2021 and stable diffusion and Midjourney were both created in 2022 and they have improved much since their beginnings.