Strauss Zelnick, the Take-Two Interactive CEO, has spent years listening to gaming's most excitable prophets announce that the next shiny thing will change everything. In 2026, that shiny thing is artificial intelligence, and Zelnick is still the guy in the back of the room with his arms crossed—not because he hates shiny things, but because he has watched a few of them rust. He was calm about Xbox Game Pass, he did not declare NFTs the future of commerce five minutes after learning what they were, and he has now applied the same unimpressed eyebrow to AI. In a 2026 industry that seems to sprint from one breathless buzzword to the next, that is practically a public service.

strauss-zelnick-says-ai-is-a-tool-not-a-game-making-genie-image-0

During an investors call, Zelnick began with a line that deserves to be printed on a mug: 'You know I'm the first person to be skeptical of other people's hype. And I would like to note that AI stands for artificial intelligence and there is no such thing as artificial intelligence.' It is a wonderfully grumpy bit of semantics, like a man refusing to call a hot dog a sandwich because he has read the dictionary and has time. Zelnick does not think AI is nothing. He simply thinks the hopes and fears surrounding it are often inflated into a parade balloon that looks impressive until someone lets the air out.

His reasoning is cheerfully analog. The handheld calculator did not stop children from learning math, so a writing bot like ChatGPT will not mean the end of essays. A machine that can generate a thousand bland paragraphs is not the same as a person who has something to say. AI hype, in his view, is a funhouse mirror: it makes the reflection enormous and slightly wrong. The tool can be useful. The reflection is not the reality.

Zelnick is especially unmoved by the idea that Take-Two's studios—Rockstar, Firaxis, Cloud Chamber, and Hangar 13—are about to be replaced by a bot that spits out games. 'And no,' he said, '[AI is] not going to allow someone to say, Please develop the competitor to Grand Theft Auto that's better than Grand Theft Auto, and then they will just send it out and ship it digitally and then that will be that. People will try, but that won't happen.' It is a lovely image: a mysterious machine in a basement, asked to out-Rockstar Rockstar, producing instead a spreadsheet with a horse on it. Ambition is not a prompt. Craft is not a command line.

But here is the twist: Zelnick is not anti-AI. He thinks AI research is a big deal, and he is interested in using it to make games. He just rejects the fantasy that the main result will be cheaper development. Instead, AI tools will 'just raise the bar' for the industry. 'We are ushering in a very exciting era of new tools,' he said, 'and they're going to allow our teams and our competitors' teams to do very interesting things more efficiently, so we're going to want to do more. We're going to want to be even more creative.'

That is a crucial distinction. AI will not shrink the ambition of game development; it will inflate it. If a team can prototype faster, animate more smoothly, upscale textures, or train smarter opponents, the market will not reward them with a shorter workweek. It will reward them with higher expectations. Machine learning is a pet octopus: fascinating, capable of opening jars, and genuinely useful around the kitchen, but not a replacement for the chef. The chef just gets a stranger, more talented assistant and a menu that suddenly includes things no one expected.

The applications are already wider than the loudest AI image generators. Machine learning helps with image upscaling, self-driving car navigation, animation prototyping, and AI opponents. Ubisoft has used a machine learning-powered animation tool called Anything World for prototyping. Google trained an AI to play StarCraft. Rocket League players recently discovered that AI can also open new doors for cheaters. Every interesting advancement tends to arrive with a catch tucked under its arm, like a party guest who brings excellent snacks and a mysterious smell.

Zelnick approaches all of this like a veteran sailor eyeing a new navigation gadget: useful, maybe, but not a replacement for knowing the sea. He does not think we are watching the end of big development teams. He thinks we are on the fringe of a new era—one in which those teams can do more. The question is not whether AI can make games. It is who gets to steer the octopus, who gets the credit, and who gets blamed when the tentacles knock over the china.

What Zelnick thinks AI will do What he thinks it will not do
Raise the bar for creativity Replace Rockstar with a prompt
Make teams more efficient Eliminate the need for human judgment
Handle tedious and technical problems Justify every breathless headline
Generate inspiration and assets Ship a better GTA on command

In 2026, the games industry is still wrestling with AI-generated art controversies, voice actor concerns, layoffs, and a thousand LinkedIn posts about prompt engineering. Zelnick's position remains refreshingly boring in the best way: subscriptions are a small part of the business, NFTs did not become the future of commerce, and AI is a tool. A powerful tool, yes. A fascinating tool, certainly. But not a magic wand, and not a genie that grants wishes to whoever types the longest prompt.

If he is right, the next era of games will not be written by a chatbot. It will be written by people with better tools, bigger ambitions, and a healthy suspicion of anyone selling unconditional enthusiasm. And for the record, the eternal search terms remain GTA 6, GTA 5 mods, GTA 5 cheats, GTA 6 cars, and San Andreas cheats—none of which, Zelnick would likely argue, can be willed into existence by a machine that does not know what it is like to want something.