Microsoft just made a move that looks contradictory on the surface — and that’s exactly why it’s worth paying attention to. The company brought in a wave of AI executives to run its Xbox division while simultaneously killing the one AI feature that was already built into the console. If you’re trying to make sense of where AI actually fits in the gaming business, this is your case study.
Xbox is under real financial pressure. Hardware revenue is down sharply year-over-year, overall gaming revenue has declined multiple quarters in a row, and the Activision Blizzard acquisition still hasn’t delivered the turnaround Microsoft was counting on. The restructuring now underway isn’t just a personnel refresh — it’s a bet on a different kind of AI strategy than the one Xbox had been pursuing.
A New CEO Rebuilds the Room With People She Trusts
Asha Sharma stepped into the Xbox CEO role in February 2026, succeeding Phil Spencer after his 38-year run at Microsoft. Sharma came directly from Microsoft’s CoreAI division, where she served as president of product — the group responsible for Azure AI Foundry, Azure OpenAI Service, and the developer tooling that powers AI-enabled applications across Microsoft’s enterprise ecosystem.
She wasted little time reshaping the leadership table. At least four colleagues from the CoreAI organization followed her into Xbox in senior roles. Jared Palmer, formerly a vice president of product at CoreAI and a senior VP at GitHub, joined to work across product, engineering, developer tools, and infrastructure. Tim Allen, who led design at both CoreAI and GitHub, took over as Xbox’s design lead. Jonathan McKay — who previously ran growth for ChatGPT at OpenAI before joining CoreAI — became Xbox’s head of growth. Evan Chaki, a CoreAI general manager, was put in charge of a team of forward-deployed engineers. David Schloss, a growth executive from Instacart where Sharma herself had worked, took the reins on Xbox’s subscription and cloud business.
This isn’t a random mix of résumés. It’s a deliberate pattern: executives who build AI-powered developer platforms, design systems at scale, and drive subscription growth through product-led growth models. Sharma is building a leadership team shaped by how AI actually ships in enterprise products, not how it gets demoed at gaming conferences.
Copilot on Xbox Is Gone — and the Reason Tells You Everything
At the same time Sharma was announcing this new leadership lineup, Xbox confirmed it is winding down Copilot on the console entirely. The Copilot assistant — which debuted on Xbox in late 2024 as a voice-activated AI that could help players navigate menus, surface game recommendations, and manage party chats — is done. Development has reportedly stopped, with servers expected to wind down by the end of 2026.
The official framing was that the feature didn’t solve a clear enough problem to justify the ongoing resource cost. That’s a polite way of saying it didn’t work well enough for gamers to want it. Internal data reportedly showed that most Xbox Series X|S owners had disabled the feature entirely, mostly because they saw no meaningful benefit and didn’t want the system overhead.
There’s a harder business lesson embedded in that data point. Copilot on Xbox was an example of AI applied as a feature — layered on top of an existing experience because the technology existed, not because the user asked for it. Gamers are not enterprise customers hunting for workflow automation. They want a responsive, immersive experience with zero friction. A voice assistant that competes for RAM and produces middling results is friction, not a feature.
Sharma’s public comments have reinforced this framing. She has been explicit that Xbox will not ship AI for its own sake, describing what she wants to avoid as “soulless AI slop” — a direct signal that her definition of useful AI is narrower and more demanding than what Xbox had been building.
What This Restructuring Actually Signals for the Gaming Industry
The apparent contradiction — hire AI executives, kill the AI feature — only looks like a contradiction if you think AI in gaming means a chatbot on your controller. Sharma’s moves suggest a different thesis: AI’s real role in gaming is in the infrastructure, the personalization engine, the subscription retention logic, and the developer tooling — not the user-facing assistant layer.
Her leadership hires align precisely with that thesis. Growth executives who ran AI-powered product loops at OpenAI and Meta, subscription strategists who understand cloud service economics, forward-deployed engineers who sit close to real technical problems — these are the people who build the kind of AI that works invisibly. The kind that recommends the right game at the right price point, reduces churn in Game Pass, and helps developers build faster on Xbox platforms.
For entrepreneurs and operators watching this from the outside, the lesson generalizes cleanly. AI does not automatically improve a product simply by being present in it. Consumer-facing AI features face a brutal test: they have to be noticeably better than the experience without them, immediately, with no learning curve. If they’re not, users turn them off or churn. Infrastructure-level AI, by contrast, can deliver compounding value without asking the user to change their behavior at all.
Microsoft is a long way from declaring victory in the gaming space. Revenue is still declining, layoffs are expected, and the competitive landscape with Sony and Nintendo has not shifted. But the leadership model Sharma is installing — AI expertise applied to subscription economics, developer ecosystems, and growth infrastructure — is a more coherent business bet than shipping a console chatbot and hoping players find it useful. Whether it moves the numbers is the next question. The strategic logic, at least, is sound.