Large technology companies' investments in AI will take years to generate returns. Before that happens, companies must first reach break-even — which will also take years.

Goldman Sachs strategist Ryan Hammond has calculated that hyperscalers like Amazon (AMZN), Oracle (ORCL), and Microsoft (MSFT) will need to generate approximately $300 billion in AI-related revenue over the coming years to cover their AI investments and achieve break-even. Hammond stated that hyperscalers still have a long way to go before reaching this target.

Hyperscalers' cloud business revenue has accelerated significantly this year. In the second quarter of 2026, their annualized revenue came in roughly $70 billion above the pre-AI-wave trendline. The group's disclosed backlog of committed orders exceeds $1.5 trillion. Large-scale AI investment is proceeding at full speed.

Hammond wrote in a research note: "We estimate that AI application-side users will need to spend approximately $1 trillion annually on AI application procurement for hyperscalers to earn meaningful returns on investment, while application-layer companies can achieve substantial profit margins on top of computing costs."

In recent weeks, driven by AI optimism, investors have once again aggressively bought hyperscaler stocks, and the risks of investing have risen accordingly. The Roundhill "Magnificent Seven" ETF (MAGS), which tracks the performance of leading AI hyperscalers, has risen 8% in one month, while the S&P 500 index posted relatively modest gains over the same period.