A string of high-profile departures from OpenAI is drawing attention to a period of transition inside one of the world’s most closely watched AI companies. From senior leadership to key researchers, exits over the past two years suggest a broader reset in priorities as the company sharpens its focus on core products and long-term strategy.
The latest to join that list is Srinivas Narayanan, a senior engineering leader who recently announced his decision to step down after nearly three years. In a post on X, he wrote, “After 3 incredible years, I am leaving OpenAI at the end of next week,” adding that he had informed leadership earlier in the month.
Sharing a longer note with his team, Narayanan reflected on the pace and scale of his time at the company. “I have decided to leave OpenAI. The last three years have been an incredible journey that felt more like ten,” he wrote. He pointed to the timing of recent and upcoming product launches as a natural moment to move on. “Leading the b2b engineering team has been an enormous privilege. With the recent/upcoming product launches, this felt like the right time to step back.”
He also looked back at how much the company had evolved since his early days. “I will also fondly remember my prior role leading the Applied Engineering team, from when it was ~40 people on a single floor in the 575 office, when I first started.”
Narayanan was closely involved in scaling some of OpenAI’s most widely used offerings. “We shipped some of the fastest-growing products in history, like ChatGPT and the API, with no real playbook to guide us,” he wrote, crediting the team behind those efforts. “This was only possible because of the incredible team we built – you are the most passionate, dedicated, and hard-working colleagues I have ever worked with. You all have inspired me so much, and I’m so proud of what we have built together. I can’t thank you enough!”
He also acknowledged the company’s leadership, including CEO Sam Altman. “I am so grateful to @sama @gdb @fidjissimo and the rest of the OpenAI leadership for this opportunity of a lifetime,” he wrote. Looking ahead, he added, “I am looking forward to spending some much-needed time with my aging parents in India before deciding what’s next.”
READ: OpenAI CEO Sam Altman sued by his sister again over sexual abuse and rape allegations (
Narayanan’s exit comes alongside several other recent departures that underline a shift in direction. Executives like Kevin Weil and Bill Peebles have stepped away as initiatives such as “OpenAI for Science” and the Sora video-generation project were either scaled back or folded into other efforts. Joanne Jang, who played a key role in building systems like GPT-4 and DALL·E, has also left after several years.
The turnover is not limited to recent months. Since 2024, a number of prominent figures have exited the company, including former CTO Mira Murati and co-founder Ilya Sutskever. Other key contributors such as John Schulman, Bob McGrew, and Barret Zoph have also moved on, while policy and safety voices like Miles Brundage stepped away citing broader concerns about the direction of advanced AI development.
At the same time, several researchers behind major model advancements, including Jason Wei and Jiahui Yu, have left for competing firms, highlighting intensifying competition for top AI talent across the industry.
As OpenAI moves from rapid experimentation toward scaling enterprise products and commercial applications, leadership and team structures are evolving in parallel. While churn at this scale can raise questions, it also reflects the pressures of operating at the forefront of a fast-moving and highly competitive AI landscape.