NextFin News — Global smartwatch shipments fell 4 percent year on year in the second quarter of 2026, according to Counterpoint Research, the first decline in a year. Fitness bands dropped further. Learning tablets and smart speakers followed the same pattern after brief rebounds tied to large-language-model features. Outside smartphones and PCs, the AI hardware narrative that dominated industry conferences has produced only temporary volume lifts followed by renewed contraction.

The pattern appeared clearly at IFA in Berlin. Chinese makers of cleaning robots, Xiaomi’s sprawling product display and the smart-home exhibits of Midea and Haier all leaned on artificial-intelligence capabilities. Few banners shouted “AI,” yet nearly every category carried some form of generative or agentic feature. The technology has become table stakes. It has not become a reliable growth engine.

Earlier optimism rested on two supports that have weakened. First, government subsidies in China expanded to more appliance categories in 2025 and temporarily boosted cleaner and smart-device sales. As those incentives tapered, volumes reversed. Second, manufacturers raised average selling prices by embedding AI functions and marketing them as upgrades. Learning tablets illustrate the shift: online average prices climbed steadily from 2023 levels into the mid-3,000-yuan range during the 2026 mid-year promotions, even as unit sales contracted.

Consumers have largely declined to treat the higher prices as justified. Many own devices already labeled “AI-enabled” yet rarely open the relevant features. Smart speakers gained large-model capabilities and health or companionship modes, yet online sales in China fell more than 40 percent in the first quarter and halved again in April. Wearables and educational tablets show the same gap between marketing language and daily use. Even successful products such as Meta’s Ray-Ban smart glasses succeed more as fashion items than as AI platforms.

The innovation dilemma is sharp. Superficial additions—voice prompts, basic summarization, or camera modules bolted onto existing form factors—fail to create compelling reasons to replace working devices. More radical experiments, from the Humane AI Pin to certain early AI companions and spatial computers, generated intense early attention and then collapsed under practical shortcomings of battery life, reliability or social acceptance. Hardware companies remain trapped between incremental features that do not move demand and disruptive concepts that the market rejects.

Broader spending caution compounds the problem. In the United States, Black Friday online spending rose mainly through higher average prices rather than higher unit volumes. Surveys have shown consumers reducing planned electronics budgets across income groups. Some explicitly prefer AI tools at work while resisting their presence in personal devices. In China, public discussion has turned toward the idea that ordinary users need not chase top-tier specifications, a direct reaction to AI-driven price inflation.

Younger buyers remain curious. Surveys indicate strong theoretical interest in new AI experiences, yet only a minority rush to purchase the latest hardware on release. That restraint reflects experience with earlier waves of “smart” devices that quickly became unused clutter. In a mature market, replacement demand requires clearer, repeated utility rather than another round of feature announcements.

The result is a sector that continues to ship AI-branded products while struggling to answer a basic commercial question: why should a consumer pay more, or replace a functioning device, for capabilities they neither request nor regularly use? Until hardware makers close that gap between technical possibility and everyday value, the AI revival in consumer electronics will remain more slogan than sustained recovery.