6-month stipend on evaluating robustness of AI agents safety guardrails and for running an AI spear-phishing study | grantmaking.ai