How OpenAI put itself at the centre of a $1T network of deals (www.ft.com)

A recent analysis argues that OpenAI has positioned itself at the center of a roughly $1 trillion network of commercial agreements — spanning cloud infrastructure deals, enterprise integrations, hardware purchases, and API usage — by turning its models into a ubiquitous platform rather than a single product. Key moves include deep partnerships with major cloud providers (notably Microsoft/Azure), broad API licensing to startups and incumbents, and embedding its models across consumer and enterprise software, creating strong distribution, recurring revenue, and dependency pathways for customers and partners. For the AI/ML community this matters because it shifts value from model research to platform control: access modalities (hosted APIs, fine-tuning, embeddings, inference vs. training), pricing, and safety layers now influence who can build and experiment. Technically, the model-centric platform multiplies demand for large-scale training/inference compute, specialist accelerators, and data pipelines, while encouraging closed-weight deployments and gated capabilities (RLHF, prompt engineering, safety filters). The result is faster commercialization and stronger network effects, but also greater vendor lock-in, reduced reproducibility in research, and heightened regulatory scrutiny — all of which will shape how startups, labs, and regulators approach AI development and competition going forward.