Apple, Microsoft and a growing list of manufacturers warn that soaring demand from artificial intelligence data centers is driving up the cost of phones, game consoles, computers and more.
The cost of everyday electronics is poised to rise in the months ahead, and the rapid expansion of the artificial intelligence industry is emerging as the driving force. From smartphones and laptops to game consoles and televisions, a widening group of manufacturers says the memory chips at the core of these products have become dramatically more expensive, and that the added cost will be passed along to consumers.
The warnings mark a notable shift for an industry long defined by falling prices, where each new generation of devices typically delivered more capability for the same money or less. Now, leaders across technology, automotive and retail are describing a supply crunch they say is largely beyond their control, fueled by the enormous appetite for memory among the companies racing to build the data centers that power AI.
A widening chorus of warnings
Among the most prominent to sound the alarm is Apple. Chief executive Tim Cook said this week that price increases on the company's next iPhone are "unavoidable," pointing to the rising cost of the memory and storage components Apple purchases in vast quantities. Cook characterized the situation as a matter of supply rather than corporate strategy, noting that the chips Apple depends on are being bought up by large technology firms pouring billions into AI infrastructure.
Apple has not detailed how large any increase might be, and the company will not introduce its next iPhone until the fall. That timing means consumers will not be able to gauge the full effect until later in the year, and even then the comparison will be imperfect, since new models routinely add features and capabilities absent from earlier versions.
Apple is not alone. Asha Sharma, who leads Microsoft's Xbox division, recently described what she called a "hardware component crisis" in an open letter to customers. She wrote that by February the price the company paid for console storage components had risen to more than twice the level of the previous fall, and that those costs had since doubled again. Looking ahead to the 2027 holiday season, Sharma said the company expects another substantial increase, which would push prices to more than five times what it paid only two years earlier. The same pattern, she added, is now affecting memory costs as well.
The concern extends well beyond consumer technology. Late last year, Dell signaled that it would raise prices, citing AI-related increases in memory costs. Ford, whose modern vehicles contain extensive computing systems, has warned that a shortage of DRAM, a common type of memory chip, could affect vehicle pricing. And in early June, a coalition of trade groups representing retailers, media companies and medical suppliers cautioned officials at the White House that an imbalance in the memory market could lead to significant and sustained price increases for American households.
A supply chain under strain
At the center of the issue is memory, one of the most fundamental and least visible components in modern electronics. Two types are most relevant. DRAM, or dynamic random-access memory, serves as the fast working memory that nearly every device uses to run software. NAND flash memory provides the storage that holds applications, photos and files. Both are produced by a small group of manufacturers, dominated globally by Samsung, SK Hynix and Micron.
That concentration matters. Because so few companies make the world's memory, and because they rely on the same kinds of factories and the same finite supply of silicon wafers, a surge in demand from one part of the market ripples across the entire supply chain. Memory production is also extraordinarily capital-intensive. Building a new fabrication plant can take years and billions of dollars, which means manufacturers cannot quickly expand output when demand suddenly spikes.
The geography of memory production adds another layer of complexity. The bulk of the world's supply comes from a handful of facilities concentrated in a few countries, with South Korea home to two of the three dominant producers and the United States to the third. That concentration leaves the global market sensitive to shifts in demand, investment decisions and trade policy, and it gives buyers little room to seek alternatives when supply tightens. With AI demand now competing directly with consumer needs for the same limited output, manufacturers hold considerable leverage over where their chips ultimately go.
Artificial intelligence has delivered exactly that kind of spike. Training and operating large AI models requires a specialized and highly profitable type of memory known as high-bandwidth memory, or HBM, which stacks multiple layers of DRAM into dense modules positioned next to the powerful processors that handle AI workloads. For memory makers, HBM and similar AI-grade products are far more lucrative than the conventional chips used in phones and laptops, and demand for them has been close to insatiable.
The demand stems from an unprecedented wave of investment. The largest technology companies have committed enormous sums to building out their AI capabilities, constructing sprawling data centers filled with specialized processors, each of which depends on large quantities of advanced memory. That spending has turned memory makers into some of the most sought-after suppliers in the global economy, and has given them little reason to prioritize the lower-margin chips that consumer devices need.
The consequence is a reallocation of resources. As manufacturers devote more of their capacity to HBM and other AI components, less remains for the standard DRAM and storage that everyday electronics require. Each wafer directed toward AI memory is one that does not become the ordinary memory found in a phone or a game console. The result is a tighter supply of those chips and, inevitably, higher prices.
Numbers that keep climbing
The scale of the increase has been striking. Industry tracking of memory prices has pointed to sharp year-over-year rises across several categories, with some types of DRAM reportedly climbing at a pace that has outstripped even the rising price of gold. Analysts have attributed the surge directly to the construction of AI data centers, which has accelerated as major technology companies compete to expand their computing capacity.
For device makers, memory is often one of the more expensive elements in a product's bill of materials, particularly in models that offer large amounts of storage. When the price of that component rises by a wide margin, manufacturers face a difficult choice: absorb the cost and accept thinner margins, reduce the amount of memory in their products, or pass the expense on to buyers. Increasingly, companies are signaling that the last option is the most likely.
The memory market has experienced sharp swings before. The industry is famously cyclical, lurching between periods of oversupply, when prices fall steeply, and periods of shortage, when they climb just as quickly. A notable shortage in 2017 and 2018 drove memory prices up significantly before new capacity and softening demand brought them back down. Whether the current episode follows a similar path depends in large part on how quickly manufacturers can expand production and whether demand from AI continues at its present pace.
Reasons for caution
There are grounds for treating the warnings with some skepticism. Corporate executives have a clear incentive to attribute price increases to external forces rather than to their own decisions, and not every cost rise blamed on AI necessarily originates there. The picture grows murkier for larger purchases. The price of a new vehicle, for example, is likely to be shaped far more by trade and tariff policy than by the cost of a single category of chip. And because new products typically arrive with new features, a straightforward comparison between this year's model and last year's rarely captures the full story.
Even allowing for those caveats, the broad direction of memory prices is difficult to dispute. Multiple independent measures point the same way, and the underlying cause, the extraordinary demand created by the AI build-out, shows little sign of fading in the near term.
A push for action
The mounting pressure has begun to draw a policy response. The recent appeal to the White House from retailers, media companies and medical suppliers reflects growing concern that rising memory costs could ripple through the broader economy, affecting not only gadgets but also the wide array of products and services that depend on electronics. The coalition framed the situation as an urgent imbalance in the market, warning that without intervention the burden could fall heavily on ordinary households.
Memory manufacturers, for their part, have an incentive to expand production to capture the current high prices, and any significant increase in capacity could eventually ease the strain. But because new facilities take years to come online, relief on that front is unlikely to be immediate.
What it means for shoppers
The effect may not stop at physical products. Data centers themselves rely heavily on memory, and as the cost of server-grade chips rises, the companies that run cloud platforms and online services could face higher operating expenses of their own. Over time, those costs can find their way to consumers in the form of pricier subscriptions and services, extending the reach of the squeeze well beyond the electronics aisle.
For shoppers, the practical effect will unfold gradually and unevenly. The clearest early test will come with Apple's next iPhone in the fall, though feature changes will complicate any direct comparison with current models. Game consoles, given the timelines described by Microsoft, may face the sharpest pressure heading toward the 2027 holiday season. Personal computers, televisions and other electronics are likely to feel the effect at varying speeds, depending on how much memory each requires and how manufacturers choose to respond.
The squeeze is likely to be felt most acutely in the higher-storage versions of popular devices. Buyers already pay a premium for phones, tablets and consoles that hold more data, and that premium is tied closely to the cost of the memory inside them. As component prices climb, the gap between entry-level and high-capacity models could widen, leaving consumers who want extra storage to shoulder a larger share of the increase. Some manufacturers may also choose to keep advertised starting prices steady while quietly reducing the storage offered at each tier, an approach that raises the effective cost without changing the sticker.
What is becoming clear is that the cost of the AI boom is no longer confined to the companies building it. The same surge in demand that has reshaped the technology industry is now reaching into the everyday market for the devices people carry, watch and play on. For a generation of consumers accustomed to electronics that grew cheaper over time, the coming months may mark a noticeable departure from that long-standing trend.
