ironquill.tech/board

$ cat jobs/research-resident-ai-bio-threats-and-mitigations-rand-corporation-83eb8557a4c9.json

Research Resident, AI × Bio Threats and Mitigations

rand corporation·US·United States·mid
Apply on himalayas → Get AI match score →
Job Type: Term (Fixed Term) About the Team The RAND Center on AI, Security, and Technology's mission is to understand and address emerging threats to global security, in particular biosecurity and AI security. We believe that biotechnology and frontier AI will be among the most powerful and transformative technologies of the coming decades, and that addressing the risks that accompany them is one of today's most pressing challenges. Our work focuses on understanding these threats and developing concrete solutions to counter them, from technical analysis through to engagement with the stakeholders positioned to act. RAND is an independent non-profit, but regularly engages with senior stakeholders within the USG, allied governments, and relevant industry stakeholders. About the Role We are looking for a Research Resident to help us understand and counter the ways frontier AI could lower barriers to harmful biological activity. You will assess where AI meaningfully increases misuse risk, develop safeguards and detection approaches, and turn analysis into concrete recommendations that developers, policymakers, and government partners can act on. Depending on your background and interests, your work could span areas such as: Assessing how AI accelerates scientific workflows in ways that could enable misuse, identifying concerning capabilities, and developing indicators for major capability shifts at the AI × Bio intersection Contingency planning for the point at which a frontier model materially increases the likelihood of serious biological misuse, and building the infrastructure to recognize and coordinate a response in advance. Understanding how sophisticated and state-linked actors might pursue AI-enabled biological misuse, and developing frameworks to deter and disrupt them, including work on attribution and signaling. Building methods and architectures for advanced technology providers to detect misuse before harm occurs, including identifying the signals and patte