Your next laptop will cost more because of AI.
Who pays for the AI boom if it's not the people building it?
TL;DR: Apple has raised prices across Macs, iPads, and other hardware, in some cases by $200 to $800 in Canada. The driver of the increase is a surge in memory and storage costs as AI data centres absorb more supply. Apple says these increases are “unavoidable.” Analysts expect pressure to last into 2027. If you need a device, buy it when you need it, but if yours still works, this is a good year to keep it.
A few weeks ago, a new MacBook Air cost $1,499 in Canada. Today the same laptop costs $1,799 for the same chip, the same screen, same everything. The only thing that changed is the price.
For the past 2 years, the monetary cost of building AI has largely lived in data centres, venture funding, and corporate balance sheets. Now a piece of that bill is showing up at the checkout. Let’s walk through what’s happening, why it’s happening, and what you can do about your hardware bills.
What Apple raised prices on in 2026 (the full list)
On June 25, Apple raised prices across its Macs and iPads, as well as the Apple TV, HomePod and Vision Pro. Microsoft followed within days. Apple’s stock fell about 6%, its worst day in over a year. Notably, the iPhone was left alone.
Here’s the damage in Canadian dollars:
In an interview with The Wall Street Journal, CEO Tim Cook said price increases were “unavoidable,” citing “huge” increases in the cost of memory and storage and calling the situation “unsustainable.”
Why memory chips got so expensive: the AI shortage explained
The culprit is memory, specifically DRAM (dynamic random access memory) and NAND, which sit inside nearly every phone, laptop and server, made almost entirely by three companies: Samsung, SK Hynix and Micron.
Over the past year, those 3 companies redirected most of their production toward a premium, built-for-speed version of memory that AI accelerators consume in enormous quantities, leaving less supply for consumer devices. By some estimates, AI data centres will consume roughly 70% of all memory chips produced globally this year.
Hyperscalers like Google, Microsoft, Meta, and Amazon are signing multi-year deals and making large prepayments to secure chips. Suppliers, in turn, are prioritizing those higher-margin orders.
Even Apple, one of the largest and most sophisticated buyers in the world, is feeling the squeeze. Cook described it bluntly: memory suppliers are “passing along huge price increases,” and Apple can’t fully absorb them anymore.
Which brings us to the price jump. DRAM prices rose 80-90% in a single quarter. Gartner expects combined memory and storage costs to climb roughly 130% by the end of 2026, enough to lift average PC prices about 17% and push global shipments down more than 10%. Morgan Stanley estimates that, even as new factories ramp up, there still won’t be enough memory to meet demand for consumer devices and supply could fall short by about 15%.
Gartner analysts even coined a word for it: memflation, price inflation in everyday tech driven specifically by the memory shortage.
This isn’t a usual chip-market wobble. Industry leaders are calling it structural, not cyclical. Manish Bhatia, an executive vice-president at Micron, called it “the most significant disconnect between supply and demand … that we have seen in a quarter century,” and IDC research director Nabila Popal described it as “a structural reset of the entire market.” Cyclical problems fix themselves. Structural ones create a lasting impact.
For the past 2 years, the cost of AI has been concentrated among the companies building and running models. Now those costs are spreading outward.
First, it was cloud pricing. Now it’s hardware. Apple is a useful signal because of its scale and supply chain leverage. If Apple is passing costs through, smaller manufacturers have even less room to absorb them.
When will memory and RAM prices drop?
Willy Shih, a Harvard Business School professor who has watched these cycles since the 1980s, has seen this movie before. “Anytime people show me these curves that just go to the sky with no end — that never continues forever,” he told Fortune. “This too will pass.”
It probably will. But Shih says this one is different: “We have never seen price increases like we have over the last six months.” And the people closest to the supply chain aren’t expecting fast relief. Yang Wang, a principal analyst at Counterpoint Research, says the squeeze runs “through the second half of 2027.” Micron’s CEO has said the first large wave of new factory capacity won’t come online until 2028. The factories that could end this are years from completion.
So: relief eventually, pressure for a while. We should plan around the “while,” not the “eventually.”
Even OpenAI is building its own chip to dodge the cost
For a sign of how real this is, look to the companies causing the shortage. They’re attempting to build their own way out of it.
On June 24, the day before Apple’s price hike, OpenAI revealed its first custom processor, Jalapeño. Designed with Broadcom, Jalapeño is built for inference, the costly, around-the-clock work of running models like ChatGPT after they’re trained. Broadcom’s CEO says it could cut those costs by roughly half, though that figure comes from the company selling it and hasn’t been independently verified, so hold it loosely.
The strategic logic doesn’t need verifying, though. Inference is the recurring tax on every AI product, and custom silicon is how a company like OpenAI stops paying full freight to its suppliers. So, the same demand pushing OpenAI to build its own escape hatch is the demand making the memory in your laptop more expensive. The largest players get to insulate themselves. Most buyers can’t.
I’ll bet that this won’t be the last time we’ll see powerful companies avoiding the costs of AI while the rest of us absorb them.
Should you buy a laptop now? What you can do about rising prices
You can’t fix the global chip market. But you can make good decisions around timing and configuration.
There’s a bigger move, too: understand what you’re looking at, and say so. When your team debates new hardware, or a colleague grumbles that “tech just got expensive,” you know the real reason: the costs of the AI build-out are being spread to all of us. Naming that in a meeting or a budget conversation is a small act of literacy.
Frequently asked questions
Why did Apple raise prices in 2026? Apple raised Mac and iPad prices because the memory and storage chips inside them have surged in cost. The cause is AI: data centres are buying up the bulk of the world’s memory supply, leaving less for consumer devices and driving prices up. Apple said it could no longer shield customers from the increase.
What is memflation? Memflation is the term analysts use for the rise in everyday tech prices caused specifically by the memory chip shortage. As AI demand pushes memory costs up, those costs flow into laptops, phones and other devices — inflation driven by memory, hence “memflation.”
Should I buy a laptop in 2026 or wait? If you need one, buy it now. Memory prices are forecast to keep climbing into 2027, so waiting will likely cost more, not less. If you’re buying anyway, pay for extra memory and storage up front, since you usually can’t add it later. If your current laptop works fine, keeping it another year is the cheapest option.
When will memory and RAM prices drop? Not soon. Analysts at Counterpoint expect the shortage to last through the second half of 2027, and Micron’s CEO has said major new factory capacity won’t arrive until 2028. Some easing may begin in late 2026, but meaningful relief is likely a 2027–2028 story.
Is the iPhone going to get more expensive too? Apple didn’t raise iPhone prices in this round — only Macs, iPads, the Apple TV, HomePod and Vision Pro. Whether the iPhone follows later is unconfirmed, so treat any claim that it will as speculation for now.
AI in the news
Microsoft’s New AI in education report highlights widespread adoption and increasing demand for support (Microsoft Source) Microsoft’s third annual report found 92% of students and education leaders and 88% of educators have now used AI for school, yet 77% of students and 53% of educators say they’ve had no formal training. It’s the clearest sign yet that the bottleneck has shifted from access to AI literacy, a gap working professionals are hitting too.
AILiteracyGuide.com can help - it’s a free website I built, with tips for parents, teachers, and workers on how to gain AI literacy.
EU AI Act transparency obligations: preparing for compliance by 2 August 2026 (Sidley Data Matters) From August 2, organizations must tell people when they’re interacting with an AI and label AI-generated audio, images, and video. And most companies aren’t ready. It’s a turning point for everyday trust: disclosure is becoming the baseline expectation, not a nice-to-have.
Anthropic accuses Alibaba of largest Claude distillation attack
(AI Weekly) Anthropic alleges Alibaba ran 25,000 fake accounts to pull ~29 million Claude conversations to train its own Qwen model, and took the evidence to the White House. Beyond the corporate drama, it raises a question every AI user should sit with: who actually owns the conversations we have with these tools?