AI Safety Manager

The role

The AI Safety Manager will own the safety policy work for Ash, a mental health wellness app. This person will research safety issues, consult experts, and decide how the AI should behave to ensure user safety.

What you will do

• Build and grow the policy book by turning issues into clear working policies.
• Work with lawyers, psychiatrists, psychologists, therapists, philosophers and researchers to make decisions.
• Work through hard trade-offs between conflicting goals like privacy and autonomy.
• Investigate user reports, emerging risks and new requirements to make clear decisions.
• Set priorities for safety issues to tackle.
• Turn policy into how Ash behaves by defining model training, guardrails, safe modes and banners.
• Build processes for triage, investigation, escalation and follow-through.
• Translate findings into changes to Ash and its surrounding systems.
• Build, test and improve safety systems quickly.
• Keep policy in step with the law by working with the General Counsel.

Required qualifications

• Understand AI, how people talk to it, and what controls a company needs.
• Enjoy writing policy and pulling together many sources.
• Worked professionally, ideally at a startup.
• Strong judgment to reason through difficult questions and weigh competing considerations.
• Open-mindedness to change approach based on evidence.
• Ability to handle a lot of problems at once.
• Ability to learn fast and seek input from others.
• Excellent researcher, writer and communicator.
• Ability to use AI research tools well.
• Experience building and improving something like a product, process, program or project.
• Resourceful and hands-on approach to investigation and coordination.
• Handle sensitive information with care and stay thoughtful under pressure.

Benefits

• Competitive compensation (top of personal market)
• Travel between NYC / London offices
• Free lunch and coffee in office
• Generous learning budget
• Personal therapy coverage

About the company

Slingshot AI is the team behind Ash, the first AI designed for mental health. It is a well-funded Series A company with offices in New York City and London.