AI Safety Manager

Slingshot AI Slingshot AI is the team behind Ash, the first AI designed for mental health. Our mission is to make support more accessible and help people change their lives in the ways they want. We’re building a world-class team by empowering individuals with the autonomy, flexibility, and support they need to do their best work. We dream big, iterate fast, and care deeply. If that sounds like you, we’d love to hear from you. Our team spans machine learning, product, engineering, conversational design, clinical, growth, and operations, with offices in both New York City and London. We're a well-funded Series A company, having raised $93M from Andreessen Horowitz, Radical Ventures, Forerunner Ventures, plus top-tier tech investors involved in ElevenLabs, Captions, Shopify, Plaid, Notion, Canva, Twitch, Airtable, and many others. The Role Ash is a wellness app that gives people a place to talk through stress, anxiety, relationships or a bad day. It isn’t a replacement for therapy. People find it useful for everyday issues. People bring deeply personal things to Ash, so new safety questions come up all the time, from user conversations, customer support, researchers and partners. We build our safety policy from what’s actually happening with our users and keep adding to it. We’re hiring someone to own this work full time: researching each issue, getting input from the right experts and deciding what Ash should do. Safety work at Slingshot runs across teams. You’ll report to our General Counsel, who leads our regulatory work, and work closely with our machine learning, engineering, clinical and research teams to put our policies into practice. Key responsibilities - Build and grow our policy book. Take each new issue and turn it into a clear working policy. Pull together the relevant regulation, professional guidance, clinical practice, ethics and research, then make a reasoned judgment call. - Bring in the right experts. Work with lawyers, psychiatrists, psychologists, therapists, philosophers and researchers, including our own clinical and research team and our advisors, and use their input to make practical decisions. - Work through hard trade-offs. Many safety questions pose conflicting goals, such as protecting other people against a user’s privacy, or steering someone toward a safer choice against respecting their autonomy. Help decide where we come down and set the process for making those calls. - Turn ambiguous issues into clear decisions. Investigate user reports, emerging risks and new requirements. Establish what we know, what remains uncertain and what we should do next. - Set priorities. Help the team decide which safety issues to tackle first, and give the people building safety features clear answers. - Turn policy into how Ash behaves. Each policy defines when it applies and what Ash should and shouldn’t do. Policies will feed directly into model training and into the guardrails, safe modes and banners that shape Ash’s responses, so they need to be clear and concrete, with examples of good and bad responses. - Make the response work in practice. Build clear processes for triage, investigation, escalation and follow-through. Define who does what, create the tools and documentation they need, and handle the operational details yourself. - Turn learning into product improvements. Work with engineering, machine learning, clinical and legal colleagues to translate findings into changes to Ash and the systems around it. - Build, test and improve quickly. For new systems, put a useful first version in place, learn from real cases and refine it. Match the pace and level of review to the consequences of the decision. - Keep policy in step with the law. Work with our General Counsel to update our