OpenAI is redirecting resources away from experimental research to focus on improving ChatGPT, prompting several senior staff members to leave the company. The shift reflects mounting competitive pressure from Google and Anthropic as the $500 billion company evolves from research lab into commercial enterprise.
Cristina Criddle reports for the Financial Times that vice-president of research Jerry Tworek, model policy researcher Andrea Vallone, and economist Tom Cunningham are among those who departed over the strategic changes. Ten current and former employees confirm that resources for experimental work are being reallocated to advance the large language models powering ChatGPT.
Researchers working on projects unrelated to large language models increasingly struggle to secure computing credits and resources. Teams developing Sora and DALL-E, OpenAI’s video and image generation models, felt neglected as their work was deemed less relevant to ChatGPT. The chatbot serves 800 million users.
In December, CEO Sam Altman declared a “code red” after Google’s Gemini 3 model outperformed OpenAI’s offering on independent benchmarks. “Companies are spending an unbelievable amount of money on that race, and that often requires focus,” a former employee says.
Chief research officer Mark Chen rejects the characterisation, stating that “long-term, foundational research remains central to OpenAI and continues to account for the majority of our compute and investment.”
Jenny Xiao, partner at Leonis Capital and former OpenAI researcher, believes the company’s advantage lies in user base rather than technical superiority. “They’re converting technical leadership into platform lock-in,” she says. “The moat has shifted from research to user behaviour.”
