AI Offense-Defense Dynamics Lead Researcher (AI)

TL;DR

AI Offense-Defense Dynamics Lead Researcher (AI): Leading research to decode the dynamics that determine whether AI systems threaten or protect society with an accent on quantitative system dynamics models and risk frameworks. Focus on identifying critical leverage points for policy interventions and developing methodologies to assess new AI models.

Location: Remote (Open Globally). US-based candidates may be hired as employees; candidates outside the US will be engaged as independent contractors. Note: No visa sponsorship provided.

Salary: $125,000 – $200,000 per year (for US employees)

Company

The Center for AI Risk Management & Alignment (CARMA) works to lower the risks to humanity and the biosphere from transformative AI through rigorous analysis, policy frameworks, and technical safety approaches.

What you will do

• Develop quantitative system dynamics models capturing technological, social, and institutional factors influencing AI risk.
• Design simulations to identify critical leverage points where policy interventions can shift offense-defense balances toward safety.
• Operationalize offense/defense dynamics taxonomy and develop metrics to predict AI system risks.
• Build empirical analytical frameworks using documented cases of AI misuse and beneficial deployment.
• Research the interaction between technical AI characteristics and sociotechnical contexts.
• Draft evidence-based guidance for AI governance and communicate findings via publications and conference talks.

Requirements

• M.Sc. or higher in Computer Science, Cybersecurity, AI Policy, Risk Management, or a related field.
• Demonstrated experience with complex systems modeling, risk assessment, or security analysis.
• Deep understanding of modern AI systems, including LLMs, multimodal models, and autonomous agents.
• Strong understanding of dual-use technologies and factors influencing offensive vs. defensive applications.
• Ability to develop both qualitative frameworks and quantitative semi-empirical models.
• Proven record of research publications related to technology risk, governance, or security.

Nice to have

• PhD in a relevant field.
• Expertise in hypergraph techniques or complex network analysis methods.
• Experience developing interactive tools or dashboards for risk visualization.
• Background in international relations theory as applied to technology proliferation.
• Experience with the quantification of qualitative risk factors.

Culture & Benefits

• 100% remote work environment with occasional travel.
• Direct impact on global AI safety, governance, and risk mitigation efforts.
• Inclusive organization promoting equal opportunities based on individual merit.
• Competitive salary and benefits for US-based employees.