Recruiting a new head of preparedness may be trickier for OpenAI than you might think.
The ChatGPT maker recently generated buzz online when it said the position — which pays $555,000 a year plus equity — is up for grabs. Yet some tech-industry observers say finding someone who’s qualified and willing to take it on poses a challenge.
Whoever lands it will be tasked with balancing safety concerns and the demands of CEO Sam Altman, who has shown a penchant for releasing products at an exceptionally fast clip. This year, OpenAI rolled out its Sora 2 video app, Instant Checkout for ChatGPT, new AI models, developer tools, and more advanced agent capabilities.
The head of preparedness role is “close to an impossible job,” because at times the person in it will likely need to tell Altman to slow down or that certain goals shouldn’t be met, said Maura Grossman, a research professor at the University of Waterloo’s School of Computer Science. They’ll be “rolling a rock up a steep hill,” she said.
Altman himself has even described the position as intense.
“This will be a stressful job, and you’ll jump into the deep end pretty much immediately,” he recently wrote on X.
Still, it could be a dream come true for the right individual. OpenAI has had a major impact on people’s lives, and the more than half a million dollars in base pay is in line with what AI talent can expect to earn these days.
Who might be qualified for the job
The posting for the position doesn’t list common requirements such as a college degree or a minimum number of years of work experience.
OpenAI said a person “might thrive” in the role if they have led technical teams; are comfortable making clear, high-stakes technical judgments under uncertainty; can align diverse stakeholders around safety decisions; and have deep technical expertise in machine learning, AI safety, evaluation, security, or adjacent risk domains.
OpenAI’s former head of preparedness, Aleksander Madry, moved into a new role in July 2024. He left a vacancy within the company’s Safety Systems team, which builds evaluations, safety frameworks, and safeguards for its AI models.
Madry has a background in academia, but a seasoned tech-industry executive would be a better fit going forward, said Richard Lachman, a professor of digital media at Toronto Metropolitan University. Academic types, he said, tend to be more cautious and risk-averse.
Lachman expects OpenAI to seek out someone who can protect the company’s public image regarding safety, while allowing it to continue innovating quickly and driving growth. “This is not quite a ‘yes person,’ but somebody who’s going to be on brand,” he said.
OpenAI’s approach to safety has raised concerns internally, prompting some prominent early employees, including a former head of its safety team, to resign. The company has also been sued by some people who allege it reinforces delusions and drives other harmful behavior.
In October, OpenAI acknowledged that some ChatGPT users have exhibited possible signs of mental health problems. The company said it was working with mental health experts to improve how the chatbot responds to those who show signs of psychosis or mania, self-harm or suicide, or emotional attachment.
