By Dan Gallagher

This is an online excerpt of the WSJ's free AI & Business newsletter. Get more insights each week by signing up for it here.

Can you use AI to make your own videogames? Sure, but why would you?

It's a natural question to ask in the midst of the latest AI disruption debate. A few weeks back, Google unveiled Project Genie, which it described as a "world model capable of generating diverse, interactive environments." Genie allows users to use Google's AI tools like Gemini and Nano Banana to create entire worlds with characters that can walk, fly, drive or even ride a horse, or any other creature they can imagine.

In other words, a videogame.

Google took pains to call Genie an "experimental research prototype" that is capped at sessions lasting only 60 seconds. But that did little to tame the reaction in a market where investors are quickly hitting the "sell" button on anything even theoretically vulnerable to being replaced with an AI tool. The stocks of videogame publishers like Take-Two Interactive and Roblox sold off harshly on the news, and are now down more than 20% for the year.

Unity, which makes "engine" software used in the production of console and mobile games, has lost more than half its market value since the Jan. 29 announcement. That's a much rougher ride than even many embattled software companies have experienced in that time.

Videogames run the gamut these days, from iPhone apps with simple mechanics to complex, triple-A properties like "Grand Theft Auto" and "Call of Duty." There's plenty of room for generative AI tools to make development quicker and cheaper. Blockbuster games these days can sport credit screens of developers and artists that rival a Marvel movie.

But the technical capability to create a high-end game doesn't mean it's going to be worth the effort. Established publishers with talented development staff still produce plenty of flops.

And professionally made TV shows and movies still draw sizable audiences even as technology has allowed viewers to create those things themselves. The ability to make your own game also doesn't mean that game will turn out good-or even be worth all the time and effort involved. As game analyst Joost van Dreunen put it in his own newsletter a few weeks back: "People enjoy agency, but only for the fun parts."

Like their embattled compatriots in software, though, videogame companies may have a hard time disproving the AI narrative for a while. Plus, soaring memory costs are going to hurt sales of game consoles and PC gaming rigs. This has a knock-on effect on actual game sales.

And Take-Two's planned release of its next "Grand Theft Auto" sequel in November has some other publishers in a holding pattern. They don't want to put their titles up against what looks to be a massive blockbuster.

That level of success would actually serve as a strong counter-argument to the notion that homegrown AI slop will take over the industry. Whether the stock market is ready to receive that message is another question.

Dan Gallagher is part of the Heard on the Street team. Subscribe to their free afternoon newsletter here.

This item is part of a Wall Street Journal live coverage event. The full stream can be found by searching P/WSJL (WSJ Live Coverage).

(END) Dow Jones Newswires

February 17, 2026 13:00 ET (18:00 GMT)

Copyright (c) 2026 Dow Jones & Company, Inc.