Ubisoft and the technology trap

Ubisoft's 2025-26 annual report reads less like a business update and more like a confession. A $1.98 billion loss, the deepest in the company's 40-year history, six canceled games, two shuttered studios, and a debt deal with Tencent that hands over partial ownership of Ubisoft's biggest franchise just to keep the lights on. CEO Yves Guillemot calls it a year of "decisive action." Investors might call it a reckoning.

The headcount math tells its own story. Ubisoft has cut staff from a 2022 peak of 20,665 down to 16,590, a 20 percent reduction, with 1,200 jobs gone since last year alone. None of it has moved the stock. So the company's next move is to lean even harder into what it calls "cutting-edge technology," specifically AI, as the path back to creative relevance. The trouble is this isn't a new idea for Ubisoft. It's the same idea, recycled again.

Go back through the last decade and you'll find Guillemot cheerleading virtual reality before PlayStation VR even shipped, sitting front row when Google unveiled Stadia in 2019, buying up blockchain startups like Sorare and Sky Mavis during the crypto boom, and calling the metaverse "the industrial revolution of tomorrow" in 2021. Every one of those bets fizzled. Stadia died in three years. Ubisoft's cloud-streaming rights, bought for roughly $100 million from Microsoft's Activision deal, now sit on the books written down to $36 million. The metaverse and blockchain investments simply evaporated from the conversation once the hype did.

Compare that to Take-Two's Strauss Zelnick, who spent the same years saying no. No to VR as an industry mover, no to the metaverse as anything Take-Two didn't already own through GTA Online, no to blockchain speculation, and, this year, a flat dismissal of AI generating a GTA game. Take-Two's share price has fared far better than Ubisoft's, and while there are plenty of variables at play, the pattern is hard to ignore: one company kept chasing tools, the other kept shipping games.

What's getting cut in the meantime is people, the actual expensive, hard-to-replace part of making games. Layoffs are quick and visible. A bad multi-year bet on a platform or engine is quiet and slow, but it can cost far more in the end, and it almost never gets scrutinized the way a headcount spreadsheet does. Ubisoft's report doesn't really answer how leveraging AI at scale returns the company to profitability. It just repeats a belief the company has held, and lost money on, several times before.