市场反应被形容为拥挤的巨型市值科技股交易出现「信心危机」，因估值看来偏高，投资人担心 1 则负面头条就可能引发连锁效应。Google 就是一例：它公布了优于预期的营收与获利，以及 Gemini 的进展，并在 2025 年缩小了与 OpenAI 等 AI 领先者的差距；然而，投资人聚焦于其支出计划比预估高出 $65 billion，使得股价仍走低。Amazon 过去一年因成长较弱且被视为 AI 落后者而遭批评，表示将大举投资扩张云端，股价下跌 10%；更广泛的软体情绪也急转直下，一项常见的产业基准指标出现自 2020 年疫情初期以来最差的 7 天表现。
低迷表象之下的核心问题是：Microsoft 或 ServiceNow 等既有业者的营收成长是否跟得上资本支出激增，以及 Anthropic 等 AI 原生公司是否会攫取不成比例的需求；事实上，交易员紧绷到连一则针对法律工作的 4 段 Anthropic 产品公告，都助长了应用软体类股的一篮子下跌。阴霾也被描述为外溢到相邻的风险资产，Bitcoin 下跌幅度足以抹去自 Donald Trump 开始其第 2 任总统任期以来的涨幅；相较之下，Meta 是相对例外，因投资人认为其 AI 投入正转化为加速的广告回报。至于稍后公布财报的 Nvidia，这波从 $135 billion 到 $200 billion 的资本支出计划被视为利多，因为预期其中多数基础设施将流向 AI 加速器及相关硬体；而该公司也表示预期今年销售额将超过 $500 billion，且可能主要由这 4 家美国云端领导者所驱动。
Big Tech is signaling another record year for data center and AI infrastructure buildout, with Amazon, Google (Alphabet), Microsoft, and Meta collectively reporting about $120 billion of capital expenditures in their most recent 3 months, much of it tied to data centers and hardware such as Nvidia chips. The spending outlook then escalated in a near bidding-war sequence: Meta guided up to $135 billion in capex for this year, Google followed with $185 billion, and Amazon topped it with an unprecedented $200 billion plan. Despite results that were broadly in line with or ahead of analyst estimates, several of these companies saw their share prices fall after highlighting the magnitude of the spending.
The market reaction is being framed as a “crisis of confidence” in crowded mega-cap tech trades, where valuations look stretched and investors fear that 1 negative headline could cascade. Google is an example: it reported expectation-beating revenue and profit plus Gemini traction and had narrowed the gap versus AI leaders like OpenAI in 2025, yet the stock still traded down as investors fixated on a spending plan that was $65 billion above estimates. Amazon, criticized over the last year for weaker growth and being seen as an AI laggard, said it would spend heavily to expand cloud, and shares dropped 10%; broader software sentiment also turned sharply negative, with a common industry benchmark suffering its worst 7-day period since the early pandemic in 2020.
Underneath the malaise is the question of whether revenue growth at incumbents like Microsoft or ServiceNow is keeping pace with the capex surge, and whether AI-native firms like Anthropic will capture disproportionate demand; in fact, traders were so on edge that a 4-paragraph Anthropic product announcement aimed at legal work helped sink a basket of application software stocks. The gloom is described as spilling into adjacent risk assets too, with Bitcoin falling enough to erase gains since Donald Trump began his 2nd presidential term, while Meta was a relative exception because investors viewed its AI spend as translating into accelerating advertising returns. For Nvidia, which reports later, the wave of $135 billion to $200 billion capex plans is positioned as supportive because most of that infrastructure is expected to flow toward AI accelerators and related hardware, and the company has indicated it expects to surpass $500 billion in sales this year, potentially driven largely by those 4 US cloud leaders.