How to apply for AI Offense-Defense Dynamics Lead Researcher

About Future of Life Institute

The Future of Life Institute is a nonprofit focused on reducing global catastrophic and existential risks from advanced AI. Its CARMA center works specifically to lower risks to humanity and the biosphere from transformative AI. This is a place for researchers who want their work to directly inform AI governance and risk management.

About the role

You will lead research to decode offense-defense dynamics: how specific attributes of AI technologies make them more likely to protect society or cause harm. You will build quantitative and qualitative frameworks that track how AI capabilities spread as either protective or harmful applications. The output is meant to be actionable for developers, evaluators, standards bodies, and policymakers.

A typical day

A typical day may involve reading new AI capability releases and research, refining a framework for classifying capabilities as protective or harmful, and drafting a memo or paper for developers, evaluators, standards bodies, or policymakers. You may also meet remotely with CARMA colleagues or external experts, and occasionally travel for workshops or conferences. Exact routines are not specified in the job details, so ask about team structure, meeting cadence, and deliverables during the interview.

Who Future of Life Institute is looking for

• Has led interdisciplinary research before, combining technical and policy or social science methods.
• Can design both quantitative and qualitative frameworks to analyze how AI capabilities proliferate.
• Understands AI risk management and alignment debates well enough to translate findings for developers, evaluators, standards bodies, and policymakers.
• Is comfortable working remotely and traveling occasionally for in-person work.

Tips for this application

• In your materials, directly address the core question in the job description: what determines whether an AI system or capability threatens or protects society. Show how your past work maps to that question.
• Name CARMA explicitly. Explain why its focus on grounding AI risk management in rigorous analysis and policy frameworks for AGI fits your research agenda.
• Describe a concrete framework you have built or would build to classify AI capabilities as protective or harmful, and how it would be used by developers, evaluators, standards bodies, or policymakers.
• Show that you can work across disciplines. Give one example of combining technical analysis with policy, legal, or social science methods.
• Because this is a contract role, state your availability, time zone, and willingness to travel occasionally. Do not assume full-time or permanent status.

What to cover in your cover letter

['Your specific research approach to decoding offense-defense dynamics, including methods and prior results.', 'How your work would produce actionable insights for at least two of these groups: developers, evaluators, standards bodies, policymakers.', 'Why CARMA and the Future of Life Institute are the right place for this research, referencing their focus on existential risk from advanced AI.', 'Your experience leading research, since the title is Lead Researcher and the role requires direction-setting.']

Research before applying

• Read the Future of Life Institute's published work and open letters on AI existential risk to understand their positions and framing.
• Study CARMA's stated mission and any public materials on AI risk management and alignment, especially anything about AGI policy frameworks.
• Look for prior research or reports on offense-defense dynamics in AI or related fields such as cybersecurity, biosecurity, or technology governance.
• Check who funds and partners with the Future of Life Institute and CARMA so you understand the policy and research ecosystem they operate in.

Likely interview topics

Based on the job description, expect questions about:

• How you would define and measure whether an AI capability is predominantly protective or harmful.
• Which attributes of AI technologies you think most influence offense-defense dynamics, and why.
• How you would design a mixed-methods study to track AI capability proliferation across society.
• How you would translate research findings into standards or policy recommendations for bodies like evaluators or standards organizations.
• How you would prioritize research questions when the knowledge gap is large and AI capabilities are advancing quickly.

Common mistakes to avoid

• Treating this as a standard AI ethics or policy role. The core is offense-defense dynamics and capability proliferation, not general AI ethics.
• Proposing only qualitative or only quantitative methods. The job explicitly asks for both.
• Ignoring the contract nature of the role or failing to clarify availability, time zone, and travel capacity.