OpenAI loosens its Azure ties, but the IPO question still comes down to compute costs

The revenue-sharing arrangement was also revised. Microsoft will no longer pay a revenue share to OpenAI. OpenAI’s revenue-share payments to Microsoft, however, continue through 2030 and are not tied to OpenAI’s technology progress.

The broader restructuring also keeps Microsoft deeply invested. Microsoft said it supported OpenAI’s move toward a public benefit corporation, or PBC, and recapitalisation. After that recapitalisation, Microsoft said it held an investment in OpenAI Group PBC valued at about $135 billion, equal to roughly 27% on an as-converted diluted basis.

Why this gives OpenAI more cloud leverage

For OpenAI, the strategic prize is optionality. If a customer deployment, capacity need or technical requirement is not best served only through Azure, OpenAI now has the right to use another cloud provider. Business Insider described the change as giving OpenAI room to work with Amazon or other cloud providers, though with conditions attached.

That does not mean OpenAI is likely to move its core workloads away from Azure overnight. The official terms still preserve Microsoft’s primary cloud role and Azure’s first-launch position for OpenAI products.

The practical effect is more likely to be negotiating power and operational flexibility. OpenAI can pursue capacity, pricing and enterprise-delivery options across more than one infrastructure route. Morningstar also framed the end of Azure exclusivity as giving OpenAI more flexibility and leverage as it moves toward a potential IPO.

The real test will be implementation. If OpenAI mostly keeps products and API capacity on Azure, the change is mainly a bargaining tool. If major enterprise deployments and product capacity begin to run meaningfully outside Azure, the cloud-market impact becomes much more visible.

Multi-cloud does not solve the compute bill

More cloud choice is useful, but it is not the same as cheaper AI. OpenAI still has to pay for training, inference and data-centre capacity at enormous scale. For investors, the central question is whether revenue from ChatGPT, APIs and enterprise products can grow faster than the cost of the computing infrastructure behind them.

That is why recent reporting has focused on infrastructure commitments. Implicator.ai, citing The Information, reported that OpenAI CFO Sarah Friar had doubts about whether the company could be ready for an IPO by late 2026, citing organisational and compliance work as well as more than $600 billion in five-year cloud-server infrastructure commitments.

A separate report said OpenAI had missed internal revenue and user targets while preparing for a possible listing, and said The Wall Street Journal had described its future compute commitments as about $600 billion. The same report noted OpenAI’s response that its consumer and enterprise businesses remained strong and that business demand continued to grow.

Those figures should be treated as reported estimates, not audited financial data from the official Microsoft-OpenAI announcement. The safer conclusion is narrower: the new agreement improves OpenAI’s ability to obtain compute and serve customers, but it does not by itself prove that the company’s cost structure is ready for public-market scrutiny.

The IPO path looks cleaner, but not guaranteed

The revised Microsoft relationship gives OpenAI a cleaner story to tell investors: it is less dependent on a single cloud channel, but it still has Microsoft as a major technology partner and investor.

That fits with earlier reporting from Fortune, which said a preliminary OpenAI-Microsoft agreement could make OpenAI’s restructuring and eventual public listing path clearer, while noting that financial details had not yet been disclosed at the time.

But a clearer path is not the same as a confirmed timetable. OpenAI’s latest partnership announcement did not set an IPO date. And the reported concerns from Sarah Friar suggest that organisational readiness, compliance systems, governance and infrastructure commitments could still affect any listing schedule.

If OpenAI does move toward the public markets, investors will likely focus less on the headline fact that Azure is no longer exclusive and more on the harder numbers: revenue quality, customer retention, compute margins, cash burn, risk disclosures and the predictability of long-term cloud obligations.

Three things to watch next

1. How much OpenAI actually uses non-Azure clouds. The right to use other providers matters, but the market impact depends on how much product capacity and enterprise deployment really moves beyond Azure.

2. Whether revenue catches up with compute commitments. Reported infrastructure commitments around the $600 billion range will keep pressure on OpenAI to show that its revenue growth can support its expansion plans.

3. Whether governance and disclosure mature before any IPO. The PBC structure, recapitalisation and Microsoft’s investment position make the listing story more coherent, but a public offering would still require much deeper financial and risk disclosure than a partnership announcement provides.

The bottom line: OpenAI has gained flexibility, not free compute. Azure is no longer the only route to customers, which strengthens OpenAI’s supply-chain and IPO narrative. But the company’s listing timetable will still depend on whether it can turn a bigger compute footprint into sustainable, verifiable revenue.