OPENAI'S FRONTIER BUILDERS ARE REWRITING WHO GETS TO BUILD

The box labeled "enterprise software" sat on a shelf for forty years. SAP, Salesforce, Workday. Per-seat licensing. Annual contracts. Human hands on every keyboard. It was a clean, profitable loop. Then OpenAI launched Frontier on February 5, 2026, and the loop started to look fragile. ## Agents That Onboard Like Employees, Not Software OpenAI introduced Frontier as a platform that helps enterprises build, deploy, and manage AI agents that can do real work, giving those agents the same skills people need to succeed: shared context, onboarding, hands-on learning with feedback, and clear permissions and boundaries. That framing matters. OpenAI is not pitching a tool. It is pitching a coworker. Barret Zoph, OpenAI's general manager of business-to-business, told reporters: "What we're fundamentally doing is basically transitioning agents into true AI co-workers." Zoph knows the gap between demo and deployment. He rejoined OpenAI in January after departing Thinking Machines Lab, an AI startup he co-founded alongside OpenAI's former CTO, Mira Murati. The fact that he came back tells you something. HP, Intuit, Oracle, State Farm, Thermo Fisher, and Uber are among the first to adopt Frontier, and dozens of existing customers including BBVA, Cisco, and T-Mobile have already piloted Frontier's approach to power some of their most complex and valuable AI work. ## The SaaS Investors Already Know What Comes Next Here is where it gets uncomfortable for anyone holding Salesforce or Workday stock. The combined rollout of Anthropic's and OpenAI's new agentic AI systems has spooked investors in traditional big enterprise SaaS companies. The concern is that AI-native upstarts will increasingly disintermediate the relationship those big SaaS providers have with their customers. If a Frontier agent can execute sales workflows without a human ever logging into Salesforce, the per-seat licensing fees that currently power the SaaS economy could lose their justification. This is not a distant hypothetical. The architecture is already live. Frontier is compatible with agents that OpenAI has built, agents that enterprises build themselves, as well as agents from third parties like Google, Microsoft, and Anthropic. That openness is the weapon. When the platform manages all agents regardless of origin, the platform becomes the infrastructure layer everything else runs on top of. Frontier appears to be OpenAI's bid to become the "operating system of the enterprise," offering a unified platform for building agents that can navigate apps, execute workflows, and make decisions. ## Clay, Decagon, Valthos: What Native AI Companies Actually Look Like Frontier Builders are creating companies built natively on OpenAI models, where AI runs real work in the real world, deploying models that write code, serve customers, identify risk, and drive decisions across domains from go-to-market and customer experience to compliance and biodefense. Three of them reveal how wide the application surface has become. Clay is built on AI agents that autonomously research markets, analyze signals, and generate outreach strategies in real time, doing more than surfacing insights by acting on them and continuously shortening the distance between idea and execution. Think of it as a sales team that never sleeps and never asks for a raise. Decagon uses OpenAI models to power AI agents for customer support and operations that can understand inbound requests, decide on next actions, and execute workflows across tools. With OpenAI's APIs, these agents can reason through complex situations, follow business rules, and improve over time from real-world interactions. And then there is Valthos, which is either the most consequential or most unsettling application depending on your read. Valthos operates AI-driven systems that analyze biological sequences as they emerge, assess risk, and prioritize response paths in hours instead of months, using OpenAI as the reasoning layer that ties together biological models, sequence analysis, and decision pipelines. Biodefense running on a commercial AI API. That sentence deserves a full stop. ## $852 Billion and Still Losing Money Every Quarter The tension in all of this is financial and it is real. OpenAI announced it closed its record-breaking funding round at a post-money valuation of $852 billion. The round totaled $122 billion of committed capital, up from the $110 billion figure the company previously announced. Amazon agreed to invest up to $50 billion, Nvidia invested $30 billion, and SoftBank invested $30 billion. Those numbers are staggering. They are also a bet on a future that has not arrived yet. The company lost an estimated $12 billion in just one quarter from July to September 2025, with total projected losses from 2023 to 2028 potentially reaching $44 billion, and HSBC analysts have said profitability is unlikely before 2030. So OpenAI is worth $852 billion and cannot turn a profit. The resolution to that contradiction is the enterprise strategy. CFO Sarah Friar has said enterprise customers account for roughly 40% of its business and is looking to grow that share to about 50%. Frontier is not a product launch. It is the path to the P&L. The ad revenue play compounds it further. OpenAI's ad pilot generated $100 million in annual recurring revenue in under two months, and the company has told investors to expect $2.5 billion in 2026 ad revenue to grow to $11 billion in 2027 and $25 billion in 2028. That projection assumes OpenAI's products reach 2.75 billion weekly users by 2030, a number that would make it the most-used software product in history. ## McKinsey and BCG Certifying on OpenAI Is the Real Signal The consulting alliances announced on February 23, 2026 matter more than they look. OpenAI said it has formed "Frontier Alliances" with Accenture, Boston Consulting Group, Capgemini, and McKinsey and Co. These firms do not partner with products. They partner with infrastructure. When McKinsey certifies practitioners on a platform, that platform is being embedded into the operating model of every Fortune 500 client those consultants touch. McKinsey's global managing partner Bob Sternfels said: "CEOs and business leaders face unprecedented challenges in capturing value with agentic AI. To scale, they must rewire their businesses, reimagining domains and evolving how their people work. McKinsey's deep domain expertise and experience with high-impact tech transformations, infused with OpenAI's leading Frontier technology, will help clients close this gap and capture real value." Capgemini's chief strategy officer Fernando Alvarez put it bluntly: "If it was a walk in the park, OpenAI would have done it by themselves, so it's recognition that it takes a village." This is not a compliment. It is a warning that the real bottleneck is not model capability. It is change management inside organizations that have been running on the same software stack since 2005. The Frontier Builders are not waiting for those organizations to catch up. They built around them. Clay, Decagon, and Valthos did not inherit legacy infrastructure. They started from the model and built outward. That is the structural advantage that a company founded in 2023 holds over an enterprise that spent a decade implementing SAP. Companies that build AI into their infrastructure will be at a compounding advantage over the next decade. OpenAI wrote that. But the Frontier Builders are the ones proving it. The $852 billion valuation is the market agreeing.