OpenAI’s new move! Splurging $4 billion to set up a new company—staying focused on getting things done instead of just talking models.

If the past three years have been a race for AI model development, then the theme for 2026 has become quite clear: how to make AI truly usable.

On May 11th, local time, OpenAI announced the formation of a new company with an initial investment of over $4 billion (approximately 27.2 billion RMB), aimed at helping enterprises build and implement AI. OpenAI has taken an important step from models to deployment.

The newly established deployment company is controlled by OpenAI and brings together 19 investment, consulting firms, and system integrators. OpenAI stated that this is a long-term partnership led by TPG, with Anhong Capital, Bain Capital, and Brookfield as co-founding partners, and Goldman Sachs, SoftBank, among others, also as founding partners.

OpenAI believes that private equity investors have extensive experience in helping companies implement operational transformations and change management. This capability complements OpenAI and its deployment company’s expertise in technology, products, and cutting-edge AI deployment. The collaboration aims to help clients build suitable AI systems and redesign workflows.

Additionally, to rapidly grow the new company’s business, OpenAI announced in a blog that it will acquire AI consulting and engineering firm Tomoro. The acquisition is expected to be completed within the next few months. According to Tomoro’s official website, the company was founded in 2023 in partnership with OpenAI, with clients including Mattel, Red Bull, Tesco, and Virgin Atlantic Airways.

After acquiring Tomoro, the new deployment company will immediately bring in about 150 experienced on-site deployment engineers. The blog mentioned that these engineers will help clients connect OpenAI models with their data, tools, control mechanisms, and core business processes, thereby accelerating the deployment of AI systems that can operate in daily operations.

Why establish a deployment company? OpenAI’s Chief Revenue Officer, Denise Dresser, believes that AI is becoming increasingly capable of handling internal organizational tasks, but the current challenge is how to help enterprises integrate these systems into their infrastructure and workflows that support their business. The purpose of the deployment company is to bridge this gap and turn AI capabilities into real operational impact.

“The real bottleneck has never been the model itself; the problem is transitioning companies from ‘cool demos’ to ‘production environments.’” said an engineer with 12 years of backend experience. OpenAI’s move is a “smart move”—deployment, monitoring, rollback, compliance… these boring tasks are where most AI projects fail.

Compared to competitor Google, OpenAI has cutting-edge AI models but lacks application scenarios and deployment experience. Establishing a deployment company allows for complementary strengths, with partners providing practical experience in large-scale enterprise transformation.

Just a few hours after announcing the formation of the deployment company, OpenAI also launched Daybreak, an AI product for cybersecurity, integrating OpenAI models, the coding product Codex, and OpenAI’s security partners to defend networks and ensure software security.

OpenAI CEO Sam Altman stated that AI has already excelled in cybersecurity and will become even more powerful. He hopes to collaborate with as many companies as possible to help them ensure their safety. From this series of actions, it’s clear that this security product is also fundamentally aimed at clearing obstacles for large-scale AI deployment within client systems.

Recently, OpenAI’s product release pace has noticeably accelerated. It faces competition from Google and Anthropic, doubts about the commercialization speed of large models, and ongoing lawsuits involving Elon Musk. The capital market’s expectations for OpenAI are turning into pressure.

At the end of March, OpenAI announced a $12.2 billion (approximately 8B RMB) financing round, the largest in Silicon Valley’s history, with a post-money valuation of $85.2 billion (about 6.1 trillion RMB). There are reports that OpenAI may go public in 2027, with an earliest possible IPO application in the second half of 2026, and an IPO valuation potentially reaching around $1 trillion (about 6.8 trillion RMB).

To go public, it’s not enough to just tell the story of models; OpenAI urgently needs a new achievement to prove itself. Whether this deployment company can become a key force in truly integrating AI into industry remains to be seen, but at least this direction is being pursued industry-wide—the AI war has reached the “last mile,” and how it operates within enterprise systems is critical.

(Source: First Financial)