Over the past few weeks, I have been writing about the upcoming games I want to play or reliving some of the old titles that I still enjoy. Now, with the start of the new year, it is time to write a more boring piece on artificial intelligence. GiN blogger Michael Blaker has written some amazing articles about how artificial intelligence works in games. Those are some great examples, but today I am here to talk about how companies are going to take this trend into new areas, and it’s not necessarily for the betterment of either developers or players.

First, Square Enix announced that they planned to have AI do seventy percent of all their game quality assurance by 2027. This is part of a trend where the game industry seems to be heading, even though they are risking the soul of their games in order to save a few dollars. While I understand that AI has been used in pathfinding, enemy movement in strategy games, and plenty of other areas like Michael pointed out, that is not what I am concerned about.

In marketing there is an adage that if you try to appeal to everyone, you’ll appeal to no one. Video games are, first and foremost, an art form. We had that debate back in the 1990s, and the “games are art” camp clearly won. Video games have been created by talented designers, artists, and programmers in order to tell a story, provide an experience, or just to be fun. Examples of this can be seen across the generations from Doom and Quake to Super Mario Brothers and Halo to Hollow Knight: Silksong and Clair Obscur: Expedition 33.

All of those titles were created by people who loved gaming, and each one appealed to a certain market or type of gamer. Bungie wasn’t trying to pitch Halo to JRPG fans, and no one recommends Super Mario Galaxy when a player says that they love Gears of War. It’s that level of customization and targeted storytelling that makes so many titles special.

However, in our modern corporate game culture (mostly in the AAA space), the selection of what titles get made is chosen by risk-adverse businesspeople. In his book The Armchair Economist, economist Steven E. Landsburg states that people respond to incentives. These publicly traded companies and C-Suite executives have a financial incentive not to back a project that might fail. They are less concerned about giving players a new title that they really want to play, and they’re more concerned about making something that will sell regardless. Microtransactions are a great example of what happens when money incentivizes game decisions. This can also be seen in the constant remasters, remakes, sequels and clones being released. Indies and AA titles don’t have the same level of backing, but they also have far fewer constraints.

So where does AI fit into this?

AI is being marketed as a solution to “streamline processes” or to “improve efficiency” for overworked designers. What the executives see is a way to lay off their workforce in order to save money. Rather than invest in the people behind the art, they are considering going all in on trying to make more titles with as few people as possible. The assumption is that AI can make games with very minimal quality drop off from the human-designed games. But the thing is, generative AI cannot by its design create anything new. Generative AI tools just take other things that have already been made, samples them, smashes them together, and then spits out a result.

Grumpy old man argument aside, apparently the early experiments have not yielded great results. Fortune magazine recently wrote an article about EA forcing developers to use AI tools, which resulted in the AI tools causing more work for those designers.

Expedition 33 was recently stripped of two major awards when it was revealed that they used generative AI to create some early assets, didn’t like them, and patched them out for human designed assets. But even though those assets were replaced, AI is a bit of a flashpoint for many people, me included. So as much as I adore Expedition 33, losing those awards were probably the right move.

My concern is that large companies are going to force this technology into as many development stages as possible with as few humans involved. I worry that it will lead to a market oversaturated with AI-generated titles that are quickly pumped out to try and earn money. That would likely cause the quality of titles to plummet and gamer frustration to rise. In trying to make development more efficient, companies that embrace AI risk killing the market all together or at least seriously hurting it.

This is denser material than my column normally focuses on, and I appreciate you all reading to this point. I know that most technology predictions don’t age well. I hope this prediction is one of them and that I will look like a fool next year.

Until then, Happy New Year everyone! Let’s all make 2026 awesome and enjoy our games.