Gartner reports a 9.3% rise in worldwide PC shipments in late 2025, marking a recovery after two years of decline. This was driven mainly by corporate IT upgrading outdated hardware to meet Windows 11 requirements, not consumer AI demand. Omdia’s data shows a 9.2% increase in global shipments in 2025, with a 10.1% rise in Q4, and growth in notebooks (8%) and desktops (14.4%). Challenges for 2026 include supply issues in memory and storage, with costs rising 40-70%, passed on to consumers. Despite this, 57% of B2B partners expect growth in their PC business, indicating strong demand for those managing supply chain issues.
According to Gartner, worldwide spending on artificial intelligence is projected to reach approximately $2.5 trillion by 2026, marking a substantial 44% increase from the previous year. This growth is driven by ongoing investments in AI infrastructure, which alone is expected to account for $1.37 trillion of the total spending. John-David Lovelock, a distinguished vice president at Gartner, emphasizes that the success of AI adoption relies not only on financial investment but also on the maturity and self-awareness of organizations.
Why do we care?
Follow the money, and you see two very different kinds of spending behaving exactly as expected.
PCs came back because Windows 11 forced replacement; no strategy decks, no debate—just operational necessity.
AI is the opposite. The spending is real, but the certainty is not. A trillion dollars in infrastructure doesn’t mean a trillion dollars in value. It means organizations are buying capacity before they’ve proven demand—or accountability.
Treating AI like the next PC cycle assumes inevitability. It isn’t. The first downturn will expose which providers sold infrastructure and which sold outcomes.
Mature markets pay the bills. Emerging ones test judgment. Endpoint spend survives downturns. Experimental spend gets interrogated.