OpenAI engineers are 8 times more likely to join Anthropic than vice versa. The trend for DeepMind, Google’s AI division, is even more striking, with an 11:1 ratio in Anthropic’s favor.
OpenAI engineers are 8 times more likely to join Anthropic than vice versa. The trend for DeepMind, Google’s AI division, is even more striking, with an 11:1 ratio in Anthropic’s favor.
This is stated in a report by venture capital firm SignalFire, which is based on LinkedIn data, Business Insider writes .
In a fierce war for AI talent, Anthropic’s strong security and technical positioning, as well as its early-stage startup status, have helped it attract the best. The company has lured top executives, including two co-founders of OpenAI. The startup itself was formed from a group of former OpenAI employees.
According to SignalFire, Anthropic also has a high staff retention rate of 80% compared to 67% at OpenAI. DeepMind follows Anthropic with a retention rate of 78%.
Both Anthropic and OpenAI are growing rapidly. Anthropic's career page lists just over 200 positions, while OpenAI has nearly 330.
OpenAI's top executives left for Anthropic in part because of its focus on AI safety.
For example, Jan Leike left OpenAI in 2024. He was the head of the alignment group that ensures future superintelligent AI systems are aligned with human values. In his resignation letter, Leike said that “OpenAI’s culture and safety processes have taken a back seat to brilliant products.” He now holds the same position at Anthropic.
OpenAI co-founder John Schulman also left last year to join Anthropic , writing on X that he wanted to “deepen” his focus on AI consensus. But he then left the startup to join another one launched by former OpenAI CTO Mira Murati, Thinking Machines .
Schulman is not the only OpenAI co-founder to join Anthropic. AI researcher Durk Kingma, who worked at DeepMind for several years, also joined.
Anthropic has also poached several prominent DeepMind employees. The company has hired DeepMind senior research fellow Neil Houseby to set up a new office in Zurich.
It also hired research scientist Nicholas Carlini from DeepMind this spring after he spent 7 years at Google. In his blog, Carlini wrote that “the people at Anthropic actually care about the same security issues that I care about, and let me work on them.”
SignalFire also partly attributes the talent shift to the fact that Anthropic's AI assistant, Claude, has gained popularity among developers.
“Engineers often gravitate toward companies whose products they admire and use,” the report says.
In attracting talent, Anthropic's early stage can also be an advantage.
OpenAI, valued at $300 billion, has been around since 2015, while Anthropic, valued at $61.5 billion, was founded in 2021. And Google has long been a public company.
The prospect of receiving equity at an early stage of development may be more attractive to early-stage companies like Anthropic, according to Zuhair Musa, co-founder of engineer compensation platform Levels.fyi.
“People may see a lot more promise in Anthropic than in OpenAI, even though growth in both companies is quite strong,” Musa said.
