The global shortage of memory and storage chips driven by AI data centers has created unprecedented cost pressures across the technology sector. As reported by Reuters, Apple has never seen a component price increase this much, this quickly, with the company increasing MacBook Air prices to $1,299 from $1,099, MacBook Pro to $1,999 from $1,699, and iPad Air to $749 from $599. Dell raised prices roughly 15% to 20% in late 2025, with Lenovo following in January 2026, while HP, Acer, and ASUS warned of similar increases. The shortage is driven by AI data centers buying up DRAM and NAND capacity, creating what industry experts describe as an insatiable demand for memory chips.
The International Monetary Fund's chief economist Pierre-Olivier Gourinchas has issued a stark warning about AI's dual impact on inflation, as reported by Bloomberg News. The AI investment boom is "generating tremendous valuations" for companies in US stock markets and countries such as South Korea, creating a wealth effect that could add to price pressures. According to Gourinchas, there are "different channels from the AI component" - one through narrow supply chain bottlenecks and one through demand-side effects, both moving in the same direction. The wealth effect comes on top of AI-driven supply constraints already feeding into consumer inflation, with consumers feeling richer and more willing to spend on vacations, homes and other big-ticket purchases. "These demand pressures, they generate inflation," Gourinchas explained, emphasizing that the demand-side impact comes on top of supply constraints already being driven by AI investment.
Apple's decision to raise prices across parts of its Mac and iPad range represents the clearest consumer-facing example of AI-driven cost pressures, as reported by The Wall Street Journal. The company has pointed to surging memory and storage costs, driven in part by the rapid expansion of AI data centres. Microsoft Corp. announced yet another price hike for Xbox consoles, adding to the consumer-facing inflation pressures. The buildout is increasingly competing for components that were once treated as ordinary consumer-electronics inputs, creating a situation where AI is being sold as a future productivity machine but currently looks more like a giant construction project.
Goldman Sachs estimates that data centres could account for close to half of US power-demand growth through 2030, with consumer electricity prices potentially rising at a faster pace through 2026 and 2027. According to Investing.com India, data centres do not just consume chips but require power around the clock, require grid upgrades and pull skilled workers into increasingly tight regional labour markets. The local nature of grid constraints means the pain will not be evenly spread, with regions having concentrated data-centre investment feeling pressure long before national inflation data fully captures it.
The CPI's information technology category, including computers, smartphones, and peripherals, has been a persistent deflationary force for decades, but that tailwind is now slowing and possibly reversing. From a Federal Reserve perspective, higher interest rates cool demand-related inflation but have a much smaller impact on supply-side inflation driven by a shortage of AI memory chips. However, if consumers and businesses are paying more for technology products, they will have less money to spend on other products, potentially offsetting some AI memory chip inflation through reallocation of spending. The AI trade is widening from chips and cloud to power, cooling, electrical equipment and infrastructure, with consumers increasingly being asked to fund the bill.
