Former meta AI researcher says the company has a culture of fear that is spreading like cancer
A former Meta AI researcher wrote an internal email criticizing the company’s internal culture on his way, describing it as toxic, fear-operated and spreading like cancer.
Listen to the story

In short
- An outgoing meta AI researcher says that the team lacks objective and mission clarity in the team
- He says the review of the performance, the trimmed creates a culture of fear
- He compares the company’s dysfunction with metastatic cancer
On his way outside the company, a former meta researcher sent an internal email, in which he raised serious concerns about the company’s AI division, stricken with fear, confusion and laxity. Essay InformationThe Culture of the company is compared to a “metastatic cancer” that is spreading throughout the organization. Blankevort was part of the team working on the Lama model of Meta. Before exiting, he wrote a long note criticizing the leadership of Meta, stating that many employees feel lost and precious. “We are in the culture of fear,” he wrote, the employee blames the continuous performance reviews and trimmed waves to damage morale and creativity.
He said that AI Division, now more than 2,000 people are strong, suffering from direction deficiency. “Most do not enjoy being here,” he said. “And they don’t even know what our mission is.” Blancwort claimed that repeated internal conflicts and vague targets are damaging the team’s ability to innovate. “It’s not just a laxity – it is a metastatic cancer that affects the entire organization,” he said.
Their criticism comes in the same way that Meta is aggressively expanding its AI operations, such as to catch rivals such as Openai and Google Dipmind. The company has recently created a new unit called Superintendent Labs, which will focus on the construction of Artificial General Intelligence (AGI). Meta is hiring top talents throughout the industry, often with a proposal of millions of dollars annually.
Earlier this week, Bloomberg It was reported that the ruined pang, which was previously leading the Foundation Model Team of Apple, joining the new superintending division of Meta. He is expected to play an important role in Meta’s push for advanced AI systems, including a smart version of its assistant. Along with Pang, Meta also hired researchers from Openi, Anthropropic and Google, including Yuanzi Lee and Anton Bakhtin.
Hiring Spree has caused tension in the AI region. Openai’s chief researcher, Mark Chen, compared the strategy of recruiting the meta to “break someone in our house and steal something”. CEO Sam Altman also claimed that Meta had offered to sign a $ 100 million bonus to woo its team. Meta CTO Andrew Bosworth dismissed these reports, saying that such compensation was limited to rare and leading leadership roles.
Despite the high-profile higher, Blanchel’s comments show that the meta still faces deep root internal challenges. According to the information, his essay has given rise to internal reflection, and serves as a reminder that the talent alone alone may not be sufficient when the underlying culture breaks.