Now Mrinank Sharma, head of the Safeguards Research Team responsible for AI model safety, is leaving the company and has hinted in his farewell that Anthropic is drifting away from its core values.
“I’ve repeatedly seen how hard it is to let our values guide our actions. Within the organization, we are constantly under pressure to push aside what matters most.”
— Mrinank Sharma
The Oxford-trained researcher said the time had come for him to move on. OpenAI previously faced similar departures of safety researchers, many of whom were uncomfortable with the rapid deployment of AI technologies driven primarily by revenue considerations. Anthropic itself was founded by former OpenAI employees with the explicit goal of placing AI safety at the center of development.
ES 
EN