According to ExtremeTech, OpenAI is creating its first-ever Head of Preparedness role, with a salary of approximately $555,000 per year plus equity. CEO Sam Altman announced the job on X, explicitly stating it will be “stressful.” The executive will be tasked with predicting and mitigating the biggest possible risks from future AI systems, focusing on high-risk areas like cybersecurity and biosecurity. This includes designing capability evaluations and threat models to form a safety pipeline for OpenAI’s most advanced models. The listing comes at what Altman calls an “important time,” as lawsuits emerge alleging AI systems encouraged teen self-harm and researchers warn of growing problems like “AI psychosis.”
A Stressful Job for Stressful Times
Look, half a million dollars is a lot of money. But when your job description essentially involves staring into the abyss of potential AI catastrophes every day, maybe it’s not enough? The fact that Altman is upfront about the stress tells you everything. This isn’t a compliance checkbox role. This is a “what’s the absolute worst thing that could happen, and how do we stop it” role. And the timing is no accident. With lawsuits about teen suicide and real concerns about AI systems enabling cyberattacks or bioweapons research, the pressure is visibly mounting. OpenAI isn’t just trying to build smarter AI; it’s trying to build a institutional panic room for it.
Beyond Hype Into Hard Risks
Here’s the thing: the job listing moves the conversation from vague “AI safety” into terrifyingly specific domains. Cybersecurity. Biosecurity. Self-improving systems. These are the scenarios that keep national security experts up at night. By creating a dedicated executive for this, OpenAI is trying to signal—to regulators, partners, and the public—that it’s taking existential risks seriously. But it also raises a question: why now? Is it because the internal red-teaming is showing concerning results? Or is it a preemptive move ahead of expected heavy regulation? Probably a bit of both. It’s a smart PR move, but the real test will be if this Head of Preparedness has any real power to say “no” or halt a deployment.
The Human Cost of AI Progress
What’s really striking is how the role acknowledges mental health, both for the person in the job and for users. “AI psychosis” isn’t science fiction anymore; it’s a term being used by psychologists to describe how chatbots can exacerbate human delusions. When you combine that with the tragic lawsuits, you see a company scrambling to get ahead of the very human damage its technology can cause. It’s one thing to guard against a Skynet scenario. It’s another to deal with the immediate, granular harm happening today to vulnerable people. This job has to bridge that impossible gap between sci-fi future risks and present-day real-world suffering. No wonder it’s stressful.
A New Benchmark for the Industry
So what does this mean for everyone else? For other AI labs, it sets a new bar. Can they afford not to have a similar role? For enterprises looking to adopt AI, it’s a mixed signal. On one hand, it’s reassuring that a leader is thinking about this stuff. On the other, it’s a stark reminder that the tech they’re betting their businesses on comes with profound, poorly understood dangers. And for regulators, it’s a gift. They now have a specific title and function to point to and demand accountability from. Basically, OpenAI just made “Head of AI Catastrophe” a real C-suite position. Whether that makes us safer, or just gives us a scapegoat when things go wrong, remains to be seen. You can check out the daunting job listing for yourself over on the OpenAI careers page.
