Miles Brundage, OpenAI’s senior adviser for the readiness of AGI (aka human-level artificial intelligence), delivered a stark warning as he announced his departure on Wednesday: no one is prepared for artificial general intelligence, including OpenAI itself.
“Neither OpenAI nor any other frontier lab is ready [for AGI], and the world is also not ready,” wrote Brundage, who spent six years helping to shape the company’s AI safety initiatives. “To be clear, I don’t think this is a controversial statement among OpenAI’s leadership, and notably, that’s a different question from whether the company and the world are on track to be ready at the relevant time.”
His exit marks the latest in a series of high-profile departures from OpenAI’s safety teams. Jan Leike, a prominent researcher, left after claiming that “safety culture and processes have taken a backseat to shiny products.” Cofounder Ilya Sutskever also departed to launch his own AI startup focused on safe AGI development.
The dissolution of Brundage’s “AGI Readiness” team, coming just months after the company disbanded its “Superalignment” team dedicated to long-term AI risk mitigation, highlights mounting tensions between OpenAI’s original mission and its commercial ambitions. The company reportedly faces pressure to transition from a nonprofit to a for-profit public benefit corporation within two years — or risk returning funds from its recent $6.6 billion investment round. This shift toward commercialization has long concerned Brundage, who expressed reservations back in 2019 when OpenAI first established its for-profit division.
In explaining his departure, Brundage cited increasing constraints on his research and publication freedom at the high-profile company. He emphasized the need for independent voices in AI policy discussions, free from industry biases and conflicts of interest. Having advised OpenAI’s leadership on internal preparedness, he believes he can now make a greater impact on global AI governance from outside of the organization.
This departure may also reflect a deeper cultural divide within OpenAI. Many researchers joined to advance AI research and now find themselves in an increasingly product-driven environment. Internal resource allocation has become a flashpoint — reports indicate that Leike’s team was denied computing power for safety research before its eventual dissolution.
Despite these frictions, Brundage noted that OpenAI has offered to support his future work with funding, API credits, and early model access, with no strings attached.
Read the full article here
