A leaked internal memo from OpenAI’s revenue chief is reshaping how the industry thinks about enterprise AI distribution. Written on April 13, 2026, the memo from Denise Dresser, OpenAI’s Chief Revenue Officer, reveals two things at once: that demand for OpenAI through Amazon Web Services has been “frankly staggering,” and that the long-standing Microsoft partnership has “limited our ability to meet enterprises where they are.”

That is a significant admission. For years, Microsoft has been OpenAI’s most important commercial partner, embedding ChatGPT and Copilot across Azure, Microsoft 365, and Teams. But as enterprise buyers increasingly operate across multiple clouds, the exclusivity of that arrangement appears to have cost OpenAI deals.

What the Memo Actually Said

Dresser’s memo, reported by CNBC and Axios on April 13, was addressed to OpenAI’s sales and revenue teams. According to reporting, the key passage reads: “Since we announced the partnership at the end of February, inbound demand from our customers for this offering has been frankly staggering.”

The memo describes OpenAI as “firing on all cylinders” to build Amazon into a scaled distribution channel, and frames the Microsoft relationship as “foundational to our success” while also noting it has limited OpenAI’s flexibility. Many enterprise buyers, Dresser noted, prefer Amazon’s Bedrock platform, which is less restrictive in how it can be packaged and resold.

The Amazon partnership, announced in late February 2026, makes AWS a key distribution route for OpenAI’s models and agentic products, including OpenAI Frontier, the company’s new enterprise agent management platform.

Why Enterprise Buyers Care About Distribution

This story might look like inside baseball between two technology giants. But for the business leaders and IT teams actually buying AI, it matters a great deal.

Enterprise software buyers tend to consolidate spending through their preferred cloud provider. If your organization is deep in AWS, buying OpenAI through Amazon means the spend can apply to existing AWS credits, existing commercial agreements, and familiar procurement processes. That removes friction and budget headaches.

The fact that OpenAI had previously only been easy to buy through Microsoft Azure meant some AWS-first organizations were buying competitors by default, not by preference. Opening that channel changes the competitive dynamic.

What This Means for the AI Vendor Landscape

The memo’s candor about Microsoft also signals something larger: the AI vendor relationships formed in 2022 and 2023 are beginning to mature, and some of them have grown complicated.

Microsoft made a defining $10 billion bet on OpenAI and has built Copilot into its entire product line. That relationship was mutually beneficial at a time when OpenAI needed capital and distribution and Microsoft needed credibility in the AI race. But as OpenAI has grown to an annualized revenue run rate above $25 billion, the terms of that relationship have become more constraining.

Anthropic has navigated this differently. Its primary cloud relationship with Google Cloud has been more straightforward commercially, and its recent infrastructure expansion with Broadcom adds redundancy. Anthropic’s enterprise revenue now exceeds OpenAI’s on a run-rate basis, suggesting that more flexible distribution contributed to that growth.

What This Means for Business

If you are evaluating AI tools for your organization, the distribution shift matters for a few practical reasons.

First, pricing and procurement. If your organization is AWS-heavy, access to OpenAI models through Bedrock means unified billing, consolidated contracts, and likely better pricing through existing enterprise agreements.

Second, flexibility. AWS-distributed OpenAI products may come with fewer restrictions on customization and fine-tuning than Azure-delivered equivalents. For businesses building custom AI workflows, that headroom matters.

Third, competition. A more open distribution model for OpenAI means more vendors can build products on top of its models and sell to enterprise clients. That expands the ecosystem of OpenAI-powered tools available to buyers, which is generally good for buyers.

The era of AI exclusivity is giving way to AI distribution at scale. The businesses that win in this environment will be the ones that understand which tools are actually suited to their workflows, rather than just which ones their cloud provider defaults to.

Enterprise DNA put together a free field guide on exactly this: the full Claude ecosystem, Claude Code, and how to roll agents out without breaking things. Get the guide.