Jerry Tworek, vice president of research at OpenAI, has announced his departure from the company after seven years of service. In a note shared on social media platform X, Tworek stated his main reason for leaving is to pursue research that is "hard to do" at the current organization.

Key Achievements During Tenure

Tworek's contributions to OpenAI span numerous breakthrough developments in artificial intelligence research. His work encompassed several critical areas that shaped the company's technological advancement:

"I've loved my time with this team. I got to work on scaling RL on robots before it was cool, training the first coding models in the world that started the LLM coding revolution, discovering chinchilla scaling before it was called chinchilla, working on gpt4 and chatgpt and most recently building a team that established new paradigm of scaling training and inference compute we commonly call reasoning models," Tworek wrote in his departure note.

Career Timeline at OpenAI

Tworek's journey at OpenAI reflects the company's rapid evolution in AI research and development:

One of Tworek's most notable achievements includes the development of OpenAI's o1 and o3 models, which demonstrated doctorate-level performance in scientific subjects including physics, biology, and chemistry.

Context of Departure

Tworek's resignation occurs amid reports of changing research dynamics within OpenAI. According to a December 9, 2025 report from Wired, employees from the company's economic research team have cited concerns about the organization moving away from fact-based research. The report suggests OpenAI has become more restrictive regarding publication of research that may present AI in a negative economic light.

These developments coincide with OpenAI's transition from a non-profit model to a for-profit organization, along with reports indicating the company remains a loss-making entity despite significant financial obligations.

Research Impact and Legacy

Tworek emphasized his continued commitment to AI advancement despite his departure. "Together, we've made machine intelligence much more useful and reliable, and I'm a die hard ChatGPT reasoning model user," he stated in his farewell message. His work on scaling models and technical foundations of Large Language Models has contributed significantly to the current landscape of AI technology and applications.

OpenAI has emerged as the most generous tech company in terms of employee compensation, offering stock-based awards that average $1.5 million per employee according to a recent Wall Street Journal analysis. With approximately 4,000 employees, the artificial intelligence company's compensation structure significantly outpaces historical precedents set by major technology firms.

Industry-Leading Compensation Structure

The scale of OpenAI's equity awards represents a dramatic departure from traditional tech company compensation models. Equilar's comprehensive analysis reveals the stark contrast between OpenAI and its predecessors:

Across major tech IPOs over the past 25 years, stock-based compensation typically accounted for only about 6% of revenue in the year before listing, making OpenAI's approach unprecedented in its generosity.

Strategic Response to Talent Competition

OpenAI's aggressive compensation strategy directly responds to intensifying competition for elite AI talent. The pressure escalated significantly during the summer when Meta CEO Mark Zuckerberg launched a comprehensive recruitment campaign targeting rival AI laboratories. Meta's packages reached extraordinary levels, with some senior hires receiving offers worth hundreds of millions of dollars, and in exceptional cases, up to $1 billion.

This recruitment offensive proved effective, drawing more than 20 OpenAI employees to Meta, including notable figures such as ChatGPT co-creator Shengjia Zhao. The talent exodus prompted OpenAI to implement immediate countermeasures to protect its workforce and maintain its competitive advantage in artificial intelligence development.

Enhanced Retention Measures

In August, OpenAI responded with substantial one-time bonuses for research and engineering staff, with some individual payouts reaching millions of dollars. The company has also announced significant policy changes to make equity compensation more attractive:

• Elimination of the six-month waiting period before equity vesting begins
• Substantial increases in overall equity packages
• Targeted retention bonuses for critical personnel

Future Compensation Projections

Investor documents from the summer indicate OpenAI's commitment to maintaining its compensation leadership extends well into the future. The company projects stock-based compensation will increase by approximately $3 billion annually through 2030. This trajectory would make OpenAI's stock compensation reach 46% of revenue in 2025, the highest percentage among reviewed companies except for Rivian, which had no revenue in its pre-IPO year.

High-Stakes Hiring Initiatives

OpenAI's commitment to premium compensation extends to new hires, as demonstrated by recent job postings. The company has advertised a senior "Head of Preparedness" position focused on AI risk management, offering $555,000 annually plus equity. CEO Sam Altman characterized the role candidly, stating: "This will be a stressful job, and you'll jump into the deep end pretty much immediately."

The compensation strategy reflects OpenAI's recognition that maintaining leadership in artificial intelligence requires not only technological innovation but also the industry's most talented researchers and engineers. As competition for AI expertise continues to intensify, OpenAI's unprecedented compensation approach may establish new industry standards for retaining critical talent in the rapidly evolving artificial intelligence sector.