OpenAI's Chief Financial Officer, Sarah Friar, revealed on Monday that the company is accelerating its push into specialized industry sectors while adopting more competitive pricing tactics to counter open-source and open-weight model rivals. This strategic shift comes as enterprise business growth continues to quicken, with the company aiming to demonstrate clear value for corporate investments in artificial intelligence.

Speaking at the Goldman Sachs Communacopia + Technology Conference in San Francisco, Friar outlined that OpenAI is concentrating its immediate focus on chip design, life sciences, and financial services. Recognizing that enterprises increasingly demand AI systems optimized for specific tasks, the company is exploring more industry-tailored solutions. Furthermore, OpenAI is experimenting with pricing structures tied directly to the business outcomes clients achieve, rather than just basing fees on raw model usage metrics.

This expansion into vertical-specific AI and the adoption of a more aggressive pricing model underscore the intensifying competitive landscape OpenAI faces. The pressures stem not only from the continuous advancement of open-weight models coming out of China but also from rivals like Anthropic, which are fiercely vying for enterprise clients. As businesses become more explicit in demanding measurable returns on their AI expenditures, technology vendors must now prove their products deliver quantifiable commercial value.

Friar also highlighted OpenAI's internal success in using its own artificial intelligence for chip engineering. The organization utilized its proprietary models to participate in the development of the Jalapeno chip, successfully managing to "tape out" the design within nine months. This critical milestone signifies the chip's blueprint has been finalized and has been dispatched to a foundry for manufacturing.

While open-source and open-weight models are widely viewed as a low-cost alternative to the frontier systems produced by companies like OpenAI and Anthropic, Friar argues that the cost dynamics may not always favor the open approach. She pointed to recent strategic actions, including an 80% price reduction for OpenAI's economical Luna model, a move she stated led to a usage surge of approximately tenfold. Additionally, the developer-centric tool Codex has seen substantial adoption, now boasting a user base of 25 million.

The company's enterprise division is demonstrating robust growth figures. Friar disclosed that between June and July of this year, enterprise business revenue climbed by 32%, outpacing the 20% growth seen in overall annualized revenue during the same period. She noted that by mid-year, revenue derived from enterprise and consumer segments had essentially reached parity—a milestone the company anticipated reaching by the end of the year, achieving it ahead of schedule.

Friar asserted that, for many businesses, directly deploying OpenAI's cost-effective models could indeed prove cheaper than running Chinese open-source systems through cloud service providers. Citing a specific comparison, she stated, "If you deploy Luna and compare its cost with GLM 5.3 from Zhipu AI in a cloud environment, our price is lower."