Startups binge on AI while big firms sip cautiously, study shows (www.theregister.com)

A16z and fintech Mercury analyzed spending patterns from more than 200,000 commercial customers (June–August) and produced a ranking of the top 50 AI-native application-layer startups — finding that early-stage firms are aggressively embedding AI while larger incumbents remain cautious. Horizontal, company-wide productivity platforms dominate (60% of the list); OpenAI and Anthropic are the two most-used vendors, and surprisingly, “vibe coding” tools like Replit (despite recent reliability and data-incidents) plus Cursor, Lovable, and Emerent all made the cut. Among role-specific apps, 12 are designed to assist human workers while five pursue outright replacement roles (two agentic legal services, an AI IT helpdesk, AI software engineering, and an automated go‑to‑market function). The report signals a bifurcated AI market: startups building AI-first products and services that could reshape workflows, versus big firms taking incremental, conservative steps. A16z predicts more end-to-end agentic products and AI-native services, but risks loom — Gartner expects >40% of agentic projects canceled by 2027, and major financial commitments (e.g., OpenAI’s reported $300B compute deal with Oracle and Oracle’s potential ~$100B borrowing needs) highlight growth-before-profit pressure. The takeaway for practitioners and investors: rapid startup-led innovation may accelerate automation and new tooling, but technical immaturity, financial strain, and hype-driven risk could produce significant churn.