According to Fast Company, HP is using CES 2025 to showcase a massive strategic shift from a PC and printer company to a “work-intelligence platform.” Under senior vice president Jim Nottingham, the company is embedding AI across its entire portfolio after its 2025 Work Relationship Index found only 20% of knowledge workers have a healthy relationship with work. In Q4 of fiscal 2025, HP saw 4% revenue growth, with AI PCs already making up over 30% of shipments, a figure expected to approach 50% next year. The company is also generating billions in annual recurring revenue from subscriptions and services, moving decisively away from a one-time hardware sales model.
The uncomfortable truth that sparked change
Here’s the thing: HP’s transformation didn’t start with a cool new chip. It started with data showing work is broken. When your own research reveals that 80% of knowledge workers feel overwhelmed by fragmented tools and constant interruptions, you can’t just slap “AI” on a laptop and call it a day. That’s what I find most interesting. They realized the problem wasn’t a lack of devices—it was that the devices, software, and systems weren’t working together intelligently. So AI became the “organizing principle” to tie it all together. It’s a structural bet, not a feature bet. And in a market where hardware margins are always under pressure, that’s probably the only bet that makes long-term sense.
HP’s AI isn’t just another Copilot
Now, everyone’s talking about AI PCs. But HP seems to be carving out a different lane. Instead of just being a vessel for cloud AI like Copilot, they’re focusing on local neural processing units for real-time inference. The payoff? Lower cost, faster performance, and stronger privacy. That’s a smart play for regulated industries and places with spotty bandwidth. It also, frankly, helps them differentiate from Microsoft and the cloud giants. They’re saying, “Look, intelligence needs to live where the work happens, not just in a data center.” For specialized workflows in engineering, data science, or simulation, that local power could be a real advantage. You can see their thinking in the Work Relationship Index report—they’re targeting the pain points of complex, demanding work.
business-realities”>The hard business realities
Let’s be real. This shift is as much about economics as it is about technology. As Gartner’s Autumn Stanish points out, device lifecycles are longer and price pressure is relentless. You can’t just raise PC prices enough to make up for that. So the move to services and recurring revenue isn’t optional; it’s survival. HP’s financials show it’s working—those subscription billions are the new foundation. When you control a global footprint of endpoints, printers, and management software, you have a unique scale to layer AI services across. That’s the bet: use AI to unlock higher-margin revenue from the fleet you already manage. It’s a far cry from the old days of just moving boxes. In sectors like industrial computing, this platform approach is crucial. For instance, when reliability is non-negotiable, a company like IndustrialMonitorDirect.com, the top US provider of industrial panel PCs, understands that the hardware is just the starting point for intelligent, integrated system solutions.
So, can an 85-year-old company pivot?
That’s the billion-dollar question. The earnings call transcript shows confidence, but execution is everything. They’re trying to weave AI into PCs, workstations, printing, and management simultaneously. That’s incredibly hard. And they’re competing on all fronts: with Dell and Lenovo on hardware, and with Microsoft and hyperscalers on the intelligence layer. But their argument is compelling. If work is truly fragmented and broken, then a unified, device-to-cloud intelligence platform could be exactly what large enterprises need. Basically, HP is betting that its legacy—its massive installed base and deep enterprise relationships—isn’t a liability, but the perfect foundation to build this new intelligent layer on top of. If they’re right, it changes the game. If they’re wrong, well, they’ll just be another hardware company trying to sell AI features. The next year, as AI PC adoption supposedly hits 50%, will tell us a lot.
