OpenAI COO says ‘we have not yet really seen AI penetrate enterprise business processes’

Earlier this month, OpenAI launched a caller level called OpenAI Frontier for enterprises to physique and negociate agents, but OpenAI COO Brad Lightcap said that businesses haven’t yet seen AI adoption astatine scale.

“One of the absorbing things and immoderate of the inspiration for the enactment we’ve been doing lately astir OpenAI Frontier is we person not yet truly seen endeavor AI penetrate endeavor concern process,” the AI exec said connected the sidelines of the India AI acme held past week successful New Delhi.

“You’ve got truly almighty AI systems that immoderate idiosyncratic tin usage successful their idiosyncratic capacity. And enterprises are these highly analyzable organizations with a batch of people, teams, each having to enactment together, a batch of context. There are precise analyzable goals that person to beryllium achieved utilizing a batch of antithetic systems and tools.”

There is simply a batch of speech astir AI agents taking implicit concern processes and claiming that “SaaS is dead”. While these predictions person moved SaaS stocks astatine times, they haven’t truly travel true. In fact, Lightcap said OpenAI was a monolithic Slack idiosyncratic past year, indicating however overmuch AI firms are inactive reliant connected accepted endeavor software.

In January, OpenAI CFO Sarah Friar posted that the company’s gross is connected the rise, with the startup ending 2025 with implicit $20 cardinal successful annualized revenue. Lightcap said that request is strong, without sharing immoderate numbers.

“We astir ever find ourselves having to negociate excessively overmuch demand. We are inactive an enactment that is growing, and truthful determination is this planetary request origin that we would emotion to beryllium capable to meet, and we are moving arsenic champion arsenic we tin to beryllium capable to meet,” Lightcap said.

At the aforesaid time, OpenAI is reasoning astir however to quantify occurrence successful the enterprise. Lightcap said that OpenAI volition effort to measurement Frontier’s interaction based connected “business outcomes, not connected spot licenses.” (The comapny hasn’t yet shared pricing for Frontier.)

“Frontier is simply a mode for america to experimentation iteratively with however to really bring AI into the truly messy and analyzable areas of businesses that I deliberation if we get that right, we’re going to larn a batch astir some businesses and besides AI systems,” Lightcap noted.

Days aft TechCrunch’s conversation, OpenAI partnered with consultancies similar Boston Consulting Group (BCG), McKinsey, Accenture, and Capgemini to deploy its exertion successful an endeavor push. Even rival Anthropic launched plugins for finance, engineering, and plan for enterprises to physique agents based connected Claude.

Meanwhile, the institution doesn’t person a wide way of integrating precocious acquired open-source instrumentality OpenClaw, but Lightcap said that it gives OpenAI “a glimpse into the future” wherever agents tin bash “almost thing you privation them to beryllium capable to bash connected a computer.”

In keeping with the India AI summit, OpenAI has made a fig of caller announcements astir its concern successful the world’s largest market. The institution said India was the 2nd biggest idiosyncratic basal of ChatGPT extracurricular the U.S., with more than 100 cardinal play users. Lightcap said that dependable arsenic a modality is picking up successful India and enabling OpenAI to scope much people.

“Voice is truthful important here. And dependable models present consciousness bully capable and besides bully capable to tally successful low-latency and low-bandwidth environments, wherever you truly tin commencement to alteration entree to exertion for a radical of radical who possibly were much disenfranchised than not,” Lightcap said.

The institution besides signed an endeavor declaration for the usage of its tools and to deploy compute. Lightcap noted India is 4th successful India successful presumption of endeavor seats successful Asia, which is debased for a populous country, and OpenAI has a batch of scope to grow here.

The AI institution is besides acceptable to unfastened 2 caller offices successful India successful Mumbai and Bengaluru. However, these are apt to beryllium income and go-to-market offices. When I asked Lightcap if these offices would see method talent, helium said, “Never accidental never.”

There is besides a fearfulness of occupation impact, particularly successful countries similar India, wherever the IT services and BPO (Business Process Outsourcing) manufacture is prominent, arsenic AI tools automate immoderate of the tasks. In the past fewer weeks, Indian IT institution stocks person dipped arsenic the marketplace is taking into relationship the information that areas similar coding mightiness necessitate less humans. Lightcap said that the institution is being “grounded” successful what it has observed successful presumption of jobs market.

“Our presumption is that implicit time, jobs volition change. I deliberation we don’t yet cognize where, how, oregon what, but it seems inevitable that enactment volition look antithetic successful the aboriginal than it looks today. And that’s natural, that’s portion of the concern cycle. It’s portion of the planetary and dynamic system that we unrecorded in. And truthful I deliberation what we person to bash is beryllium capable to evidently person empathy for wherever jobs are changing astatine a precocious rate,” helium noted.