Sam Altman has advertised a $555,000-a-year role to tackle the most serious risks of artificial intelligence.
The new head of preparedness at OpenAI will be tasked with anticipating and mitigating threats from increasingly powerful AI systems.
The role includes protecting against risks to mental health, cybersecurity, and biological weapons.
It also involves preparing for the possibility of AI systems training themselves and acting unpredictably.
Altman warned the job would be stressful and demanding from day one.
He said the post was critical to helping the world manage AI’s rapid advance.
Industry leaders have echoed concerns about unchecked AI development.
Mustafa Suleyman said anyone not feeling some fear “is not paying attention”.
Google DeepMind co-founder Demis Hassabis has warned of AI systems going “off the rails”.
Regulation remains limited, leaving companies largely self-policing.
OpenAI recently acknowledged its models are becoming far more capable at hacking.
Rival firm Anthropic has reported early examples of AI-assisted cyber-attacks.
The new role also comes as OpenAI faces lawsuits linked to alleged harmful advice from ChatGPT.
OpenAI says it is improving safeguards and responses to signs of emotional distress.
