OpenAI has undergone notable changes as it intensifies its commitment to developing its flagship ChatGPT large language model technology. This focused strategy has coincided with the exit of several distinguished senior staff members, signaling a major realignment within the company's research and development divisions.
The prominent departures include Vice President of Research Jerry Tworek, Model Policy Researcher Andrea Vallone, and Economist Tom Cunningham. These exits were reported in reliable financial media outlets on Tuesday, emphasizing the impact of the company's shift toward LLM-centric projects.
According to Mark Chen, OpenAI's Chief Research Officer, the organization remains steadfast in its dedication to foundational, long-term research endeavors. Chen highlighted the advantage of integrating cutting-edge research with practical deployment, suggesting this combination enhances scientific rigor through accelerated feedback loops, rapid learning, and continuous refinement.
However, an internal reprioritization appears to have concentrated resources primarily on LLM development. Reports indicate that research teams working on non-language model projects, including those responsible for initiatives such as Sora and DALL-E, have faced funding reductions or outright rejection of additional financial support. Several non-language research programs have reportedly been discontinued within the context of a broader organizational restructuring aimed at streamlining efforts around core language model technology.
Despite multiple requests for comment, OpenAI has not yet provided a response to inquiries regarding these staffing and strategic shifts.
In parallel with this refocused mandate, OpenAI's CEO Sam Altman has remarked on the company's hiring approach. Rather than aggressive workforce expansion, Altman suggested the firm could significantly reduce the pace of new hires while simultaneously driving higher productivity, leveraging advancements in AI to maximize output without proportionate increases in personnel.
Previously, in December, OpenAI publicly declared a 'Code Red' situation, during which it paused monetization efforts to concentrate more intently on product quality improvements. This maneuver was in response to competitive pressures following the unveiling of Google’s Gemini 3 AI model.
Financially, OpenAI has reported revenue surpassing $20 billion in 2025, underscoring robust market demand for artificial intelligence technologies. Nonetheless, the company is known to incur substantial annual expenditures exceeding $17 billion, largely attributed to the computationally intensive nature of its operations. Subscription revenues derived from offerings like ChatGPT are presumably insufficient to cover these high costs fully. This fiscal dynamic places OpenAI under increased scrutiny to demonstrate viable revenue-generating strategies that justify its estimated valuation of around $500 billion.
To address monetization challenges, the company announced in January plans to experiment with advertising on its free and 'Go' subscription tiers. OpenAI has assured users that any advertisements will be clearly marked, physically separated from AI-generated responses, and will not influence the output generated by ChatGPT. User privacy is also emphasized as a priority, with conversations remaining confidential and paid subscription tiers staying free of advertisements.
These strategic and organizational developments highlight a critical juncture for OpenAI, as it balances the pressures of advancing sophisticated AI models, optimizing internal talent deployment, managing significant operational costs, and developing sustainable revenue streams.