For the past few years, Silicon Valley has lived in a rhythm of optimism fuelled by a vision of generative artificial intelligence. However, the beginning of 2026 brings a painful sobering. Investors, hitherto willing to put a premium on far-reaching ambitions, have begun to ask a fundamental question: when will billions of dollars of infrastructure investment start to translate into margins in real terms?
Microsoft has been hit hardest by this change in sentiment. The Redmond-based giant, which became a symbol of the AI boom thanks to its early alliance with OpenAI, has lost nearly 17% of its market value since the beginning of the year. That’s a loss of $613 billion, which has brought the company’s valuation below the $3 trillion barrier. The concern is not just the pace of spending, but the growing pressure from the Gemini model from Google and Claude agents from Anthropic, which are increasingly biting into the market position of Azure and Copilot.
We are seeing a similar pattern with Amazon. Despite the company forecasting an increase in capital spending of more than 50% year-on-year, the market has reacted with a sell-off that has depleted its capitalisation by $343 billion. Investors are sending a clear message: ’empty’ promises of future dominance are not enough. Apple, Alphabet and even legacy star Nvidia have also seen declines, suggesting that the valuation ceiling for pure potential-based projects has been reached.
At the same time, capital began to flow towards companies with more tangible operational advantages. TSMC and Samsung, which supply key physical components, gained a combined market value of more than $560 billion. What’s more, Walmart has joined this elite group, proving that in times of market uncertainty, investors are seeking refuge in stable business models that can use technology to optimise existing sales, not just build futuristic castles in the sand.
The current correction does not mark the end of AI, but its transition into a mature phase. The market is no longer rewarding mere presence in an arms race. What matters now is capital efficiency and tangible products that solve business problems in the here and now. For technology leaders, this means the end of a period of discounts and the beginning of a struggle to prove that artificial intelligence is an investment and not just an expensive experiment.
