Oracle’s Q1 FY27 earnings, released on September 11, 2026, revealed a remarkable 121% year‑over‑year jump in cloud infrastructure revenue—reaching $7.4 billion—driven by surging demand for AI compute capacity. The company delivered over 300,000 GPUs and expanded data center capacity by 850 megawatts, pushing its remaining performance obligations (RPO) to a record $664 billion, up $209 billion year‑over‑year. Oracle raised its full‑year revenue guidance to at least $90 billion, though free cash flow turned deeply negative due to aggressive capital expenditure.(huggingnews.com)
Meanwhile, OpenAI launched its Agents API in public beta on September 10, 2026. The new API provides developers with a managed Codex-based agent harness, enabling durable, long‑running sessions, automatic context compaction, tool orchestration, and multi-agent coordination without requiring developers to build these infrastructure layers themselves. The API supports MCP, custom functions, web search, parallel subagents, and flexible execution environments (OpenAI-hosted or self-hosted), and is available to all developers with no separate fee.(reddit.com)
These two developments underscore a pivotal moment in enterprise AI. Oracle’s explosive infrastructure growth reflects the insatiable demand for AI compute, while OpenAI’s Agents API lowers the barrier to building sophisticated, agentic workflows. Together, they signal a shift from experimentation to scalable deployment of AI agents across industries.