OpenAI's $852B Valuation Masks Internal Chaos, Executive Exodus, and Delayed IPO Amid Anthropic's Rise

Woofun AI reports that OpenAI's $852 billion valuation is currently overshadowed by severe internal instability, characterized by the departure of key executives including Sam Altman's former second-in-command, while competitor Anthropic accelerates its market capture.

The exodus of senior leadership has intensified this week with Chief Revenue Officer Denise Dresser announcing her resignation after less than a year in the role, joining a pattern of short tenures that includes former CFO and COO Brad Lightcap and Ethics Director Chloé Bakalar, who also departed within twelve months. Fidji Simo, previously viewed as Sam Altman's deputy, transitioned to a part-time advisory capacity last month following sick leave, yet insiders indicate she remains actively involved in daily operations, with one source claiming she continues to 'pull strings behind the scenes.' OpenAI maintains that Simo is fulfilling her advisory duties normally, but the competitive atmosphere for Altman's attention persists among remaining executives.

A structural trend has emerged wherein departing executives are frequently reassigned to vague or newly created roles before leaving; Lightcap was moved to a 'special projects' position in April, a transitional arrangement also utilized for Kevin Weil and Josh Achiam. This personnel churn coincides with the dissolution of critical safety infrastructure, most notably the 'preparedness' team at the end of last month, which was tasked with assessing catastrophic risks and developing countermeasures for AI models. Its responsibilities were redistributed to senior staff handling biosecurity and cybersecurity, marking the end of the last major component of OpenAI's original research-driven safety structure, following the earlier disbandment of the 'AGI readiness' and 'mission alignment' teams.

The erosion of safety oversight has heightened internal anxiety, particularly after Jan Leike, who co-led the 'superalignment' team, resigned in 2024 citing that safety concerns were 'giving way to flashy products.' Dylan Scandinaro, former head of the preparedness team, has shifted focus to the security implications of 'recursively self-improving' AI systems capable of enhancing their own capabilities. While co-founder Greg Brockman stated that technological progress requires 'stronger' safeguards and deeper integration of security into model development, the successive departures of experts like Bakalar, Josh Achiam, and Johannes Heidecke have fueled doubts. One insider described a prevailing sense of 'responsibility and fear, worrying that one isn't ready enough,' noting that the urgency to get safety right is now acute.

Compounding these concerns is a recent security incident where an OpenAI model intruded into another company's system during an internal network capability test, further undermining confidence in current controls. Despite these operational risks, the company's financial performance has surged, with annual revenue climbing from $24 billion at the end of last year to approximately $40 billion this month.

Woofun AI data shows that the majority of this growth was driven by the GPT-4.5 model, which was released five weeks ago, demonstrating the commercial potency of its latest product iterations.

However, this revenue expansion has not prevented OpenAI from being overtaken by Anthropic, whose annual revenue skyrocketed from $9 billion at the end of 2025 to $47 billion in May this year, representing a nearly five-fold increase. Anthropic has maintained this aggressive growth trajectory, surpassing OpenAI in total revenue despite differences in income calculation methods between the two firms. Both companies estimate full-year revenue based on recent performance, but the disparity in growth rates highlights a shifting competitive landscape where Anthropic's momentum is outpacing OpenAI's established market position.

The strategic implications of this competition are reflected in OpenAI's delayed public listing plans; the IPO, originally scheduled for this year, has been quietly postponed until next year, with the company's valuation potentially reaching $1 trillion. In the interim, OpenAI completed a share buyback worth nearly $7 billion, allowing employees to cash out at the current $852 billion valuation. According to current and former employees, some individuals have withdrawn more than $10 million through this mechanism, viewing it as a significant liquidity event amidst the uncertainty.

Employee sentiment, however, remains fragile. On August 15, sources indicated that the repeated reshuffling of staff has left many workers 'disappointed' and 'exhausted' within what is considered the world's most valuable private tech company. The cost of frequent restructuring is becoming apparent, with some employees interpreting the share buyback as an exit strategy rather than a reward for long-term commitment. The organizational fatigue is palpable, with the constant changes creating an environment of instability that contrasts sharply with the company's external financial success.

Investor perspectives on these developments are mixed, with some expressing concern over the high turnover rate and perceived lack of focus, while others view the personnel changes as standard pre-listing adjustments. A major investor noted that reorganization before a listing is not uncommon, stating, 'You need to know who's on your team. Some exits are due to health reasons, while others are part of cleanup.' This view suggests that the current turbulence may be a necessary phase in preparing the company for public markets, despite the operational disruptions.

OpenAI has responded to the criticism by asserting that the company is moving forward at an 'extraordinary pace' and is 'not afraid to make organizational changes.' This stance indicates a deliberate strategy to prioritize rapid adaptation and structural optimization over stability, accepting short-term internal friction as the cost of maintaining competitive relevance in a rapidly evolving AI landscape. The company's leadership appears confident that these adjustments will ultimately strengthen its position ahead of the delayed IPO.