In November 2025, Google announced the launch of their newest generative AI model: Gemini 3 Pro Image, more commonly known as Nano Banana Pro. It was tested to be seriously outperforming its predecessors and competitors in everything from basic reasoning to image and visual understanding. The model scored a 31.1% accuracy rating on the hailed test for artificial
intelligence, the ARC-AGI-2 benchmark. Other, well-known models like GPT-5 (ChatGPT) remained below 20%. While still far from the human average score of 60% accuracy, Nano Banana Pro proved to be the most advanced AI model around.
This new model is predicted to blur the lines between fact and fiction. Users can generate ultrarealistic images in seconds, photos that lack the typical red flags for AI. Nano Banana Pro doesn’t mess up text, hands, patterns, or even lighting, all markers that the public has relied on since gen-AI started to become integrated with media.
The only identification left with this model is the SynthID watermark. On every generated image, Google adds an invisible watermark that will be identified if someone were to run the image back through Google’s Gemini. Unfortunately, there are many drawbacks to this solution: screenshots could blur the watermark, users can successfully crop out enough of the image that the watermark is unrecognizable or a user could utilize a different AI model to create the image, each scenario returning a false negative
for AI being used. If the public grows to rely on SynthID identification, they could ignore these possibilities.
Both WS students and teachers are concerned about the growing use of AI and its unpredictability.
“[On social media], I have to check the comments,” said senior Olive Han. “When they say ‘this is AI’, I go ‘Wow it is.’ It’s insane. I think [AI] has its place. In some cases, I think it can be helpful to make study charts. That’s really nice, it takes off a lot of time for me, but there’s a line, a slippery slope.”
Many AI experts recommend that everything online be treated with caution, especially if it seems particularly inflammatory or harmful. Placing more emphasis on consuming news from reputable, fact-checked sources will help avoid spreading false or misleading information.
Han is concerned about the future effects of the technology, especially on Gen Z and the generations following.
“I think more people are noticing, but people still use it. It does worry me because this generation is going to become adults, they’re going to be in charge of so many people. I fear that this will affect how the next generation will grow up and see this and be influenced in that way,” said Han.
There are few laws or federal regulations on AI and its usage, so there is very little precedent for protection. However, federal representatives are starting to take notice of this issue. In the summer of 2025, the Senate voted to block a proposed 10-year ban on regulations of AI in states, so there may be an increase in federal action to oversee and manage this new technology.
