The global semiconductor industry is riding a once-in-a-generation wave of artificial intelligence (AI) investment, boosting chipmakers’ fortunes to new highs. However, analysts at Morningstar caution that this AI-powered rally may be obscuring mounting risks, warning that the sector’s notorious boom-bust cycle is far from being tamed and could resurface as early as 2026.

The AI Surge: Extending the Upcycle

Over the past two years, the explosion in AI adoption—spurred by advances in large language models like OpenAI’s ChatGPT—drastically increased demand for cutting-edge semiconductors. Market leaders such as Nvidia, TSMC (Taiwan Semiconductor Manufacturing Company), and Samsung have reported record financial results, with Nvidia’s revenue surging over 250% year-over-year during multiple quarters in 2024 and 2025. Globally, semiconductor industry revenues are forecast to top $630 billion in 2025, up from $470 billion in 2021, according to Gartner estimates.

This unprecedented demand is fueled by hyperscalers—tech giants like Amazon Web Services, Microsoft Azure, and Google Cloud—as they race to build the infrastructure powering generative AI, from massive GPU clusters to advanced memory modules. AI server and accelerator capital expenditure rose by nearly 40% in 2024, outpacing overall IT infrastructure growth. Consequently, foundries—especially TSMC—are running at record utilization, benefitting from multi-year orders and government incentives, including over $50 billion in U.S. CHIPS Act funding.

Cyclical Risks Resurface

Despite the euphoria, Morningstar analysts urge caution. In a report released September 16, they warn, “Foundries and memory makers are exposed to the intense cyclicality of the semiconductor sector.” The typical semiconductor cycle lasts around four years. AI investments have only prolonged the current rally, not broken the mold.

Recent industry data reveal early warning signs. Semiconductor billings growth, a bellwether of the sector’s health, decelerated markedly in mid-2025. The Semiconductor Industry Association (SIA) reported monthly global chip sales rising just 2.6% in July 2025 on an annual basis, compared to over 18% growth in the same period of 2024.

Much of the sector’s incremental growth outside of AI has stagnated. Consumer electronics—the former engine of semiconductor demand—remains weak, with smartphone and personal computer sales struggling to recover to pre-pandemic levels. According to IDC, global smartphone shipments shrank 7.2% in Q2 2025 year-over-year amid inflationary pressures and muted replacement cycles.

AI Spending: Nearing Its Peak?

Morningstar forecasts AI-related chip spending will crest in 2025 before slowing down as macroeconomic headwinds and market saturation set in. “The market is too upbeat on long-term AI investment growth,” the report states. Risks of a cyclical downturn in 2026 or 2027 are rising as the pace of AI infrastructure build-out normalizes and consumer demand for non-AI chips stays muted.

While state-of-the-art AI chips—such as NVIDIA’s H100 and upcoming Blackwell series—remain in tight supply and drive historic profit margins, older memory and logic segments are already seeing price pressure and inventory buildups. Major suppliers like Micron and SK Hynix have warned of volatile DRAM and NAND pricing in recent earnings calls.

Foundries’ Relative Resilience—But Not Immunity

Within the supply chain, pure-play foundries like TSMC appear better positioned. Their technological superiority—TSMC began high-volume 3nm chip production in 2024—and major U.S. investments, such as the $40 billion Arizona fab, provide buffers against sudden demand swings. Still, Morningstar emphasizes that “even leaders like TSMC remain exposed to the cyclical swings that regularly sweep through the sector.”

Chipmakers are further challenged by geopolitical uncertainty. Tensions between the U.S. and China, and ongoing restrictions on advanced chip exports, have led firms to rethink global supply chains and invest in onshoring and regional diversification.

Investor Caution: Disconnect Between AI Hype and Financial Reality

Investor enthusiasm is at an all-time high, with a record proportion of S&P 500 firms mentioning AI during Q2 2025 earnings calls, according to Goldman Sachs. Yet, the same research found only a limited number of companies are able to quantify meaningful AI-driven boosts to their current bottom lines, as most investments remain in the infrastructure buildout phase. This raises questions about how soon, and by how much, AI will measurably enhance corporate profits and productivity.

Furthermore, Goldman Sachs notes that AI’s economic impact may be understated in official data, since semiconductor and AI infrastructure investments are often classified as intermediate inputs and not fully captured in GDP figures, masking their transformative effects at a macroeconomic level.

Looking Ahead: Boom, Bust, or a New Normal?

As the cycle matures, the semiconductor industry faces both opportunity and danger. AI remains a powerful secular driver, but fading consumer electronics demand, persistent inflation, and potential macroeconomic shocks could quickly shift market sentiment. Chipmakers are investing heavily in R&D and capacity to weather future downturns, with capital spending expected to exceed $200 billion globally in 2025.

Morningstar’s outlook is a timely reminder: even in a high-tech gold rush, the basic economics of supply, demand, and cyclical fluctuation have not been erased. Industry watchers should prepare for potential volatility, even as the long-term prospects for AI and advanced semiconductors stay bright.