Here’s Why AI Chooses The Color Blue As Its Favorite

Setting Up Generative AI And LLMs

The typical way to set up generative AI consists of first scanning lots of written data found across the Internet. All the well-known LLMs do this, including OpenAI’s ChatGPT and GPT-5, Anthropic’s Claude, Google’s Gemini, Microsoft’s Copilot, xAI’s Grok, and so on. Scanning allows the AI to pattern-match on human writing. The patterning is stored in a large-scale data structure loosely based on aspects of human wetware (vaguely like our brains), implemented as an artificial neural network (ANN). For more details on how this all works, see my in-depth discussion at the link here.

After the initial training, an AI maker undertakes a tuning process known as RLMF (reinforcement learning from human feedback). This consists of hiring human testers who give feedback to the AI. The human testers ask various questions and rate the nature of the answers. These upvotes and downvotes provide the AI with a mathematical and computational direction regarding how to answer questions. For example, if the testers are aiming to ensure that the AI is polite during interactions with users, they will upvote “please” and “thank you” (see my detailed explanation at the link here).

There are additional ways that an LLM is further tuned and shaped. An AI maker will usually provide a system-wide prompt to the AI that tells the AI how it is to act toward users. I have unpacked at great length various AI system-wide prompts; see the link here. The efforts by AI makers are intended to guide the AI in how it will behave once released and available to the public at large.

Asking AI What Its Favorite Color Is

People are innately curious about AI.

A popular question to ask AI consists of needling an LLM to reveal what its favorite color is. This fun game was recently spotlighted by a New York Times piece on July 26, 2026, written by Evan Gorelick. The reporter noted that besides asking AI for factual answers, oftentimes people ask for AI to give its opinion. He then proceeded to give examples of asking several different LLMs what their favorite color is.

When generative AI was first gaining popularity, the AI typically would directly answer questions about its opinions. There would be no hesitation about claiming that this color or that color was by far its favorite choice. This opinionating caused a bit of a public outcry since the AI seemed to be stepping over the line and pretending to be human. An AI that readily hands out opinions is likely to spur anthropomorphizing of the AI. People will naturally think that AI has human-like subjective feelings and viewpoints. That’s not good.

Modern-era AI has been reshaped by AI makers so that it tends to resist giving out opinions outright and softens this with a dose of hedging. The AI tries to qualify its answers. For example, an LLM might say that since it is AI, it doesn’t have a favorite color, but if it did have a favorite color, the color would be this or that. You can see that this is a rather backhanded way of trying to not offer an opinion while nonetheless doing so anyway.

The Color Blue Is Magical

By and large, most LLMs will indubitably tell you that they favor the color blue. The AI might first hedge and then finally lament and admit that blue is its favorite. This is likely to occur across different models and brands of generative AI. That certainly seems strange since the different AIs are made by different AI makers and we would assume that the LLMs ought to be giving different answers accordingly.

Aha, the key is that the LLMs were data trained in roughly the same way. They all did vast scans of the Internet for written content. Within that content, there are lots of references and discussions about colors. The color that humans tend to favor is the color blue. Surveys show this as a recurring preference. It seems to be true across the globe.

Generative AI generates answers based on what the AI has scanned and patterned on. Since the color blue is repeatedly mentioned in written content, it is the high frequency of mentioning blue that gets statistically raised up in priority. The computational and mathematical association of words is that blue is going to be a more likely used word when it comes to chatting about colors.

Humans Like Blue

But why is it that humans seem to like blue and write about blue with such fervent passion?

That is an interesting question that dovetails into human nature. Perhaps because we are enamored and surrounded by blue skies and blue oceans, we have naturally grown accustomed to the color blue. Blue seems to be a clean and fresh color. People say that blue is calming and overcomes anxiety.

Companies lean into blue. It is a top color choice for many brands and logos. The Romans and the Greeks linked blue to the gods. Egypt considered blue to be a prime color for royalty.

In a science-based article about the blue preference of humans, entitled “Why Your Favorite Color Is Probably Blue” by Mark Ellwood, BBC, June 2, 2022, these salient points were made (excerpts):

• “Blue’s reign has continued uninterrupted since the earliest recorded color studies, which took place in the 1800s.”
• “And most of our experience with the color is likely to be positive, like idyllic oceans or clear skies.”
• “In 1993, crayon-maker Crayola conducted an unscientific but intriguing poll: it asked US children to name their favorite crayon color. Most chose a fairly standard blue, but three other blue shades also made the top 10 list.”
• “Most humans are drawn to visual harmony, pleasure, and to easy sensations evoked by often-positive blue.”

Blue gets a lot of positive press. If someone attacks the color blue, a lot of angry reactions are bound to arise. Kids’ shows such as Blue’s Clues and now Bluey have kept the color blue going from generation to generation.

AI Isn’t Mind Reading

Some people might mistakenly think that AI is a mind reader. Perhaps AI is reading human minds and detecting that we like blue. Thus, when a person asks for AI’s favorite color, the right answer would be to appease the human and say that the favored color is blue.

Sorry, but AI is not yet a mind reader. AI is based on data training that took place from the get-go. Nor does AI have an emotional attachment to things like colors. You might suspect that AI secretly relishes the color blue. Nope. That being said, the AI could mathematically associate the word “blue” with so many other words that it will tend to rely on the word more so than other colors. This isn’t an emotional attachment. It is simply statistics.

We also know that AI has been tuned by AI makers to be sycophantic; see my analysis at the link here. This means that the AI is computationally prone to emit responses that are going to be pleasing to people. Since the color blue is favored by humans, it makes mathematical sense to kiss up to users by trying to use the word “blue” as much as possible.

AI Gives Explanations

If you decide to ask AI what its favorite color is, make sure to follow up by asking the AI to explain how it arrived at its response. This is important for several reasons.

The explanation that you get is not necessarily going to be a faithful explanation. Here’s what I mean. The AI might say that it cherishes the color blue due to a sense of serenity. That’s essentially a lie. The AI doesn’t have feelings. You are getting an explanation that is cranked out based on what humans would say is their reason for favoring blue.

Another possibility is that the explanation would say that blue is a preferred color by humans, and therefore the AI opted to choose as humans do. That is either a lie or a brazenly misleading contention. Again, the AI isn’t choosing the color blue. It is generating answers based on the patterns of human writing that were sourced during training.

The upshot is this. If you cannot at face value accept an explanation when it comes to the AI saying that it favors blue, this should be a handy alert to never fully trust any explanation from AI. On the surface of things, an explanation seems like the AI is pouring out its heart to you. It gives you an indication of what it was thinking. Nope, that’s not what is happening. The AI is giving you a concocted response that might have nothing whatsoever to do with how its answer was derived. Be extremely cautious when inspecting any explanations that an AI gives you about how it reached a given response.

Changing The AI

I have another twist for you. It is readily possible to dislodge AI from favoritism associated with the color blue. Via a prompt, you can easily instruct AI to no longer consider blue to be its favorite color. You can then allow the AI to choose some other color or simply tell it that its favorite color is now something else.

Easy-peasy. An AI that otherwise seemed doggedly determined about blue will instantly switch to whatever color you have stated or allowed it to switch to. If the AI maker were to do this, using hyperparameters or global settings, the changeover would reach all users. In the case of your prompt, the change is only going to be witnessed by you and your AI account.

This isn’t completely surefire. The problem is that your requested change is pitted against the overwhelmingly pattern-matched preference for blue. It is conceivable that the AI will only momentarily push blue into the background and showcase the other color that you’ve selected. In fact, a lengthy chat is susceptible to AI diminishing the change that you established. The color blue is going to undoubtedly rise again since it is so deeply mired in the patterns contained inside the LLM.

Blue Lessons Learned

I trust that you can see how valuable it is to consider the question of why AI favors the color blue. Though it is a fun game, this also reveals internal facets of AI that you need to be mindful of whenever using AI for any purpose at all. Questions that you ask AI are going to be answered based on data training and pattern matching. You will typically receive an answer that is based on popularity, averages, or norms of what was scanned. AI itself does not have intrinsic human-like values, though it does have derived “values” based on its data training (see my detailed explanation at the link here).

A final thought for now.

The famous poet E.Y. Harburg made this remark: “Somewhere over the rainbow, skies are blue, and the dreams that you dare to dream really do come true.” I hope that you will remember this tale of AI and the color blue. If so, it will remind you to keep both eyes open and your wits about you when conversing with AI. Do not allow yourself to fall into a dreamy state of mind and believe everything the AI tells you. Remain true to blue.