OpenAI’s CFO Sarah Friar told employees on Wednesday that the company “will be a public company in 2027” — and possibly sooner if the business keeps growing. The announcement, made at an internal all-hands meeting and first reported by CNBC, marks the clearest public signal yet on OpenAI’s IPO timeline, as the AI industry moves toward a new era of publicly traded AI companies.

“It is not a finish line, it is a milestone, another fundraise,” Friar said of the planned debut. She noted the company raised $122 billion in March, framing the IPO as one more capital event rather than an endpoint for the organisation.

When asked about rival Anthropic potentially listing first — with reports suggesting Anthropic could file publicly as soon as September — Friar was direct: it’s fine. OpenAI is running its own race. No rush to beat a competitor to the bell.

Where OpenAI Stands

The backdrop for this IPO announcement is a company that’s grown into something that barely resembles what it was two years ago. OpenAI’s models now serve more than 1 billion active users weekly. The company confidentially filed its IPO prospectus with the SEC in June, setting the legal wheels in motion without committing to a firm debut date.

The valuation attached to that filing has been reported at around $1 trillion, which would make it one of the largest technology IPOs in history if the company lists at that level.

Recent pricing moves also tell a story. In late July, OpenAI cut GPT-5.6 Luna pricing by 80%, down to $0.20 per million input tokens, citing efficiency improvements. Those kinds of cuts look different when you’re preparing to face quarterly earnings calls and shareholder scrutiny. They look like a company building usage numbers that justify a massive market cap while also boxing out competition on price.

What This Means for Businesses Using AI

For business leaders who rely on OpenAI’s APIs or build products on top of their models, this IPO trajectory matters in practical ways.

Pricing pressure will continue. OpenAI in the public markets will need to show user growth and revenue expansion simultaneously. That typically means continuing to cut prices on older models while launching higher-margin premium tiers. The trajectory for API costs is almost certainly downward over the next 12 months, which is good news for teams building AI-powered products at scale.

Enterprise agreements will become more structured. Public companies face different accountability around enterprise contracts. Expect more formal SLAs, clearer pricing schedules, and better uptime commitments as OpenAI prepares to explain its enterprise business to institutional investors. That’s actually good for enterprise buyers who’ve been operating on terms that could change at short notice.

The competitive window is compressing. Anthropic potentially IPOing in September, followed by OpenAI in 2027, means both companies face pressure to capture enterprise market share before listing. That’s happening right now through pricing, features, and partnerships. Enterprise teams sitting on AI adoption decisions should recognise that the window for advantageous commercial terms from these companies may be shorter than it looks.

Vendor diversification becomes more important. A public OpenAI answers to shareholders, not just its mission. For businesses that have built significant operations on GPT APIs, understanding the difference between OpenAI’s commercial and research priorities matters more. Maintaining relationships with multiple AI vendors, including open-weight alternatives, is sensible risk management as these companies mature.

The Bigger Picture

Two AI companies racing toward billion-dollar IPOs signals that the industry is no longer in a startup phase. These are major financial events, and the companies best positioned to benefit are those that are actually deploying AI in workflows and operations today rather than waiting for the technology to settle.

For organisations deploying AI agent workforces and voice employees, this market maturation is broadly good news. More competition at the foundation model layer means cheaper and better inference over time. The differentiation that matters is the business logic built on top of that infrastructure, the workflows, the integrations, the domain-specific knowledge that generic models cannot replicate.

The race to public markets is OpenAI and Anthropic’s story. The race to operational advantage built on top of that infrastructure is yours.

---

Enterprise DNA helps organisations build AI-powered operations with Omni by Enterprise DNA — AI agent workforces, voice employees, and custom AI applications that run on whichever foundation models make sense for your business.