For a long time, AI images have had something “off” about them. While your parents and grandparents may have been fooled by Shrimp Jesus, many younger internet users who have spent time with these programs know what to look for to determine “real” from “fake.” For example, some AI images have a slight yellow tint, others appear too smooth, and in general, there’s just a vibe of “not real” to them that’s difficult to put in easily explicable terms.

At least, there was. Recently, users have been posting images generated by Google’s new Nano Banana Pro image generator, and the general response from the internet has been “We’re screwed.”

Nano Banana vs Nano Banana Pro

We’re cooked. pic.twitter.com/LRoJhALaZD

Advertisement

— sid (@immasiddx) November 24, 2025

Are there still *some* telltale signs of AI generation if one looks closely? Sure — text can appear garbled, errors can be found in repeating patterns, and proportions can occasionally appear “off.” That said, the new generation of AI image creators is really, really good. So good, in fact, that they’re definitely going to be a problem in terms of “believing literally anything you see online” in the future.

For their part, Google has tools that allow one to tell whether an image has been generated using its new, powerful AI. That said, they’re not tagged as AI-generated by default — meaning that every image you see online has to be manually checked to ensure that it’s real.

Good luck explaining this to your family at every upcoming get-together!