AI's economic boost isn't showing up in GDP, and Goldman says that's a $115 billion blind spot (www.businessinsider.com)

Goldman Sachs says the economic boost from AI is materially bigger than official statistics suggest: its analysts estimate AI has lifted US real activity by about $160 billion (0.7% of GDP) since 2022, but only roughly $45 billion (0.2% of GDP) is reflected in government accounts — leaving a roughly $115 billion “blind spot.” The bank points to a huge rise in revenue at AI-infrastructure firms (~$400 billion since 2022) and estimates about $75 billion spent on cloud-based model development and enterprise AI that isn’t captured in investment figures or corporate earnings disclosures. The gap arises from how the Commerce Department’s Bureau of Economic Analysis treats high‑performance semiconductors and related hardware as intermediate inputs: imported chips are subtracted from GDP and their value doesn’t get capitalized when used to build AI systems. That classification, plus one‑time import frontloading ahead of 2025 tariffs, understates normal AI investment and hides the creation of intangible assets (trained models and AI platforms) whose output value isn’t fully measured. For the AI/ML community and policymakers this matters — it implies national accounts and corporate reporting undercount productivity gains and investment in AI, which can distort policy decisions, R&D incentives, and how we assess AI’s true economic footprint.