€1.3B Loss Turns Far Cry 7 Into Ubisoft's AI Gamble

The real question is not whether Ubisoft can make NPCs talk more; it is whether a publisher nursing a €1.3 billion operating loss can afford to turn unfinished generative AI into part of its creative strategy.

Is Far Cry 7 an AI feature plan, or a financial stress test?

Ubisoft is reportedly using an early build of the unannounced Far Cry 7 to test generative AI tools while the company is under severe financial pressure, according to Notebookcheck. That framing matters. This is not a confirmed Far Cry 7 feature reveal. It is an R&D signal from a company trying to prove it still has a path to scale.

The claim began with Tom Henderson of Insider Gaming, who briefly posted on X that the generative AI work currently “looks like sh*t” before deleting the post. Henderson later clarified that Far Cry 7 is being used for research and development, meaning fans should not assume AI-generated content will appear in the final game.

That distinction protects Ubisoft from one conclusion but sharpens another. If this is not yet productized, then Ubisoft is testing generative AI because it wants the option. The company’s earnings language points in the same direction: Ubisoft said it is increasing investments in generative AI for dynamic NPCs and quality assurance bots.

Henderson said Far Cry 7 is being used for R&D, not necessarily as proof that AI-generated content will ship in the final release.

MLXIO analysis: the risk is that Ubisoft’s AI bet reads less like creative confidence and more like financial urgency. A publisher can experiment. A distressed publisher experimenting inside a major franchise invites a harsher question: is the technology improving the game, or is the game becoming a test rig for the technology?

Why does a €1.3 billion operating loss make AI more tempting?

Ubisoft’s latest financial picture explains the pressure. The company reported a record €1.3 billion operating loss for the 2025-26 fiscal year. Net bookings fell by over 17 percent to €1.53 billion. Ubisoft canceled seven projects, delayed six more, and cut roughly 1,200 employees over the past year.

A €1.16 billion cash injection from a Tencent transaction helped stabilize the balance sheet, but Ubisoft also warned investors that the next fiscal year will be a low point for free cash flow before any projected rebound.

That is the environment in which AI becomes attractive to management. Not because it has already proved it can make better games. Because it offers a narrative: faster iteration, cheaper testing, more dynamic content, and possibly fewer bottlenecks in expensive production pipelines.

Ubisoft’s stated AI targets are narrower than the hype cycle around the technology:

The company’s Teammates project is especially revealing. Built on Google Gemini and led by Ubisoft’s La Forge division, it aims to create NPC companions that can remember player interactions and respond in real time. Ubisoft has also shown related AI technology with Nvidia and Inworld AI at the Game Developers Conference.

MLXIO analysis: this is where the tension sits. AI can be a tool for prototyping and testing. But if the business case drifts toward replacing craft with generated volume, the result may be more content and less design. That would be particularly dangerous for a company already trying to regain trust around major franchises, a pressure we also examined in Black Flag Remake Turns Ubisoft’s Promise Into a Trap.

Can conversational NPCs make Far Cry 7 better, or just noisier?

The most optimistic version of Ubisoft’s plan is straightforward: NPCs stop repeating canned lines and start reacting to player behavior with more context. In theory, conversational companions could remember past interactions, respond in real time, and make a world feel less static.

But generative dialogue is not automatically good game design. More speech can create more problems. NPCs need tone control. They need mission clarity. They need to stay within the fiction. They need to avoid contradicting the game’s authored structure. They need to support pacing rather than sprawl across it.

That is why Henderson’s reported assessment matters. If early tests “look like sh*t,” the issue may not be just output quality. It may be integration quality. A chatbot can produce sentences. A game needs those sentences to land at the right moment, with the right intent, inside a system that still feels authored.

Ubisoft’s own positioning suggests it knows this. The company has not announced generative AI as a Far Cry 7 selling point. Henderson’s clarification also separates Far Cry 7 R&D from Teammates, which Ubisoft has discussed more openly as an experimental playable AI project.

MLXIO analysis: Far Cry 7 is useful to Ubisoft as a sandbox precisely because it is valuable. Testing inside a major franchise environment tells the company more than testing inside a sterile prototype. It also raises the stakes if the work leaks before it is ready.

Why does this feel familiar after Ubisoft Quartz?

Ubisoft has chased controversial new tech before. In late 2021, the company launched Quartz, an NFT platform tied to Ghost Recon Breakpoint. That effort was later abandoned after intense player backlash.

The comparison is not that NFTs and generative AI are the same technology. They are not. The comparison is strategic. In both cases, Ubisoft moved toward a heavily marketed tech category while players questioned whether it improved the game experience.

The market reaction to Ubisoft’s AI language has already been rough. When the company announced accelerated AI investments in January, the news triggered a 34 percent one-day collapse in its share price and pushed its market cap below €1 billion, according to the source material.

That does not prove investors rejected AI alone. The wider financial context was already ugly. But it does show that “we are investing in AI” is not a magic repair phrase when the core business is under strain.

Related MLXIO reading on production and launch execution includes Early Access Loses Hours as 007 First Light Dumps Preload, which sits in the same broader question publishers face: how much friction players will tolerate before a release even arrives.

Who benefits if Ubisoft’s AI works, and who pays if it does not?

For players, the best case is cleaner reactivity. Less repetition. NPCs that remember what happened. Interactions that feel less detached from the player’s choices.

The downside is just as clear. If AI dialogue feels generic, inconsistent, or visibly automated, it can break immersion faster than silence would have. A bad line written by a human can be edited. A bad generative system can keep producing new failures at scale.

For developers, Ubisoft’s framing leaves room for useful tooling. QA bots could support testing. R&D systems could help prototype interactions. Teammates could teach Ubisoft where conversational AI fits and where it does not.

The danger is production pressure. When a company has canceled projects, delayed others, and cut roughly 1,200 employees, every new efficiency tool will be judged against headcount and cost. Ubisoft has not said these AI investments are meant to replace workers. But the financial context makes that concern harder to dismiss.

For executives and investors, AI offers a turnaround story. Ubisoft can tell the market it is investing in systems that may reduce friction and create new forms of player engagement. That story only works if the output is strong enough to survive contact with players.

For readers tracking AI productization outside games, MLXIO’s coverage of OpenAI Codex Stops Making iPhone Users Babysit Tasks shows a separate but relevant dynamic: AI tools gain credibility when they remove visible friction, not when they demand more supervision from users.

Will Far Cry 7 become Ubisoft’s AI showcase or its warning label?

The safest path for Ubisoft is restraint. If early results remain weak, generative AI should stay in R&D, QA, or tightly bounded interactions rather than core story content. That is not a retreat. It is product discipline.

The stronger evidence to watch would be specific: Ubisoft showing NPC interactions that remember player behavior, stay consistent across sessions, and improve mission flow without producing obvious filler. Evidence against the thesis would be equally clear: vague AI branding, leaked low-quality outputs, or a final product where generated content feels detached from authored design.

Ubisoft also confirmed that new entries in Far Cry, Assassin’s Creed, and Ghost Recon are planned before March 2029, according to the related source material. That gives the company a long runway for AI experiments, but not unlimited patience from players or investors.

MLXIO analysis: Far Cry 7 may not ship with generative AI at all. Yet the reported testing already says something important about Ubisoft’s direction. The company is not treating AI as a side experiment. It is testing it against the kind of franchise environment that could define whether generative NPCs become a premium feature — or another expensive buzzword attached to a bruised publisher trying to buy time.

The Bottom Line

• Ubisoft’s reported AI testing raises questions about whether generative tools are being used to improve games or reduce production pressure.
• A €1.3 billion operating loss makes cost-saving technology more strategically important for the publisher.
• Using a major franchise like Far Cry for AI R&D could shape player trust if the technology feels unfinished or intrusive.