Lenovo is building an AI assistant that ‘can act on your behalf’

While most of the focus in the AI ​​race is on model builders and cloud platforms, Lenovo is closer to millions of users than most companies. As the world’s largest PC maker by volume, Lenovo ships tens of millions of devices each year. What it decides to deliver, package and integrate can directly shape how AI appears in many daily lives.

That’s what made Lenovo’s announcement today at CES remarkable. At a flashy event Tuesday at Sphere in Las Vegas, he introduced Qira, a system-level, cross-device AI assistant designed to run on Lenovo laptops and Motorola phones. It’s Lenovo’s most ambitious AI effort to date and a rare glimpse into how a global hardware giant plans to more deeply integrate AI.

Jeff Snow, head of AI products at Lenovo, talked to me about how Qira came to be, why the company deliberately avoids a single exclusive AI partnership, and what he’s learned from past experiences like Moto AI and Microsoft’s recall debacle.

Qira emerged from a quiet but significant internal reorganization less than a year ago, according to Snow. Lenovo has removed AI teams from individual hardware units such as PCs, tablets and phones and centralized them in a new software-focused group that works across the company.

For a company long optimized around hardware SKUs and supply chains, this move marks a shift toward a more prominent place for AI. “We wanted integrated, cross-device intelligence that works with you throughout the day, learns from your interactions and can take action on your behalf,” Snow said. He mentioned using Qira’s built-in template on his flight to CES to help him learn how to talk about current events in meetings based on the notes and documents on his PC.

Qira is not built around a single flagship AI model. Instead, it’s modular. Under the hood, it mixes local on-device models with cloud-based models, anchored by Microsoft and OpenAI infrastructure accessible through Azure. Stability AI’s delivery model is also integrated, as well as links to application-specific partners like Notion and Perplexity.

“We didn’t want to limit ourselves to just one model,” Snow said. “This space is moving too fast. Different tasks require different trade-offs in performance, quality and cost.”

This stance flies in the face of pressure from large AI labs, many of which would happily become the proprietary intelligence layer for a company of Lenovo’s reach. Lenovo’s view is that optionality is more important, especially given its control over one of the world’s largest consumer computing distribution channels.

Snow previously worked on Moto AI, Motorola’s assistant, which he said saw high initial engagement. More than half of Motorola users have tried it, but retention hasn’t been good. He said too much of the experience felt like prompt-based chat features that people could already get elsewhere.

“It took us away from competing with chatbots,” Snow said. “Qira is about the things that chatbots can’t do, like continuity, context, and direct action on your device.”

Lenovo also paid close attention to the backlash around Microsoft’s recall feature. Snow said Qira is designed from the ground up with opt-in memory, persistent flags and clear user controls. Context ingestion is optional. The recording is visible. Nothing is collected in silence.

Cost pressures weigh on all of this. Memory prices are rising as demand for AI strains supply chains, and analysts expect PC prices to follow. Qira doesn’t increase the base system requirements for PCs, Snow said, but it works better on high-end machines with more RAM. Lenovo is working to shrink local models to smaller memory footprints, like 16GB of RAM, without watering down the experience.

Strategically, Lenovo views Qira as both a means of retention and a safeguard against hardware commodification. In the short term, the company hopes that closer integration between laptops and phones will entice customers to stay within the Lenovo ecosystem. Longer term, Snow introduced Qira as a way to differentiate Lenovo devices when specs alone are no longer enough.