Policy Design Manager, Conventional Weapons

📍 Job Overview

Job Title: Policy Design Manager, Conventional Weapons

Company: Anthropic

Location: Remote-Friendly (Travel-Required) | San Francisco, CA | New York City, NY | Washington, DC

Job Type: Full-time

Category: AI Safety & Policy Operations

Date Posted: 2026-08-22

Experience Level: Mid-Senior Level (5-10 years implied)

Remote Status: Hybrid (25% in-office expectation, with potential for more)

🚀 Role Summary

• Own and maintain Anthropic's conventional weapons policy, establishing clear boundaries for AI model usage in this sensitive domain.

• Develop and implement robust threat models and evaluation systems to assess and mitigate risks associated with AI contributing to weapons development, particularly in software and autonomy components.

• Collaborate closely with engineering teams to translate policy into practical model guardrails, advanced detection systems, and effective enforcement tooling.

• Act as the primary subject-matter expert for escalations related to conventional weapons content, ensuring rapid and informed responses to emerging threats.

• Foster a comprehensive understanding of the policy across product, engineering, legal, and leadership teams through clear communication to both technical and non-technical audiences.

• Engage with external experts, government bodies, and industry partners to refine policy and enforcement strategies, ensuring Anthropic's AI systems remain safe and beneficial.

📝 Enhancement Note: This role is deeply embedded within the "Safeguards" organization, focusing on the operationalization of AI safety policies. The "conventional weapons" aspect implies a highly specialized and critical area of AI risk management, demanding a blend of technical understanding, policy expertise, and risk assessment skills. The role's emphasis on "dual-use" technologies suggests a need for nuanced judgment and the ability to define boundaries where clear lines are not readily apparent.

📈 Primary Responsibilities

• Policy Ownership & Maintenance: Lead the development, refinement, and ongoing maintenance of Anthropic's conventional weapons policy, ensuring it accurately reflects evolving technological landscapes and misuse vectors.

• Threat Modeling & Evaluation: Design and execute comprehensive threat models and rigorous evaluations to identify and quantify risks of AI models contributing to weapons development, encompassing software, autonomy, and related components.

• Guardrail & Detection System Development: Partner with AI/ML engineers and software developers to translate policy into actionable model guardrails, robust detection mechanisms, and effective enforcement tooling.

• Subject Matter Expertise & Escalation Management: Serve as the go-to expert for complex escalations concerning conventional weapons content, providing rapid analysis and actionable recommendations.

• Cross-Functional Communication & Alignment: Develop and deliver clear, concise communications about the conventional weapons policy and its rationale to diverse audiences, including technical teams, legal counsel, product managers, and executive leadership.

• External Engagement & Policy Influence: Proactively engage with external subject matter experts, government agencies, and industry stakeholders to gather insights, build consensus, and inform policy development and enforcement strategies.

• Risk Assessment & Mitigation: Continuously assess emerging risks and misuse patterns related to conventional weapons and AI, adapting policies and systems proactively.

• Operationalization of Policy: Ensure policies are not just theoretical but are effectively integrated into the AI development lifecycle and operational systems.

📝 Enhancement Note: The responsibilities highlight a proactive and operational approach to AI safety. The emphasis on "turning policy into model guardrails, detection systems, and enforcement tooling" signifies a hands-on engagement with the engineering and product development lifecycle, characteristic of advanced AI policy and safety roles. The need to "serve as the subject-matter expert for escalations" and "rapid response to emerging risks" points to a critical, on-call component of the role.

🎓 Skills & Qualifications

Education:

• Bachelor's Degree or an equivalent combination of education, training, and/or experience.

• Required field of study: A field relevant to the role as demonstrated through coursework, training, or professional experience. Experience:

• 5-10 years of experience in a relevant setting, such as a service research laboratory, a defense research or innovation agency, or a company that designs or builds weapons systems.

• Experience translating complex technical evidence into sound policy judgments.

• Experience in operationalizing policy and developing clear, operationally precise documentation. Required Skills:

• Deep Applied Expertise in Weapons Systems: Proven understanding of conventional weapons, their components, and their development lifecycle.

• Policy Development & Articulation: Ability to write clear, operationally precise policy and explain complex technical topics to non-specialist audiences.

• Technical Analysis (Open Source): Proficiency in conducting rigorous technical analysis of weapons systems using only open-source intelligence.

• Legal Framework Understanding: Grasp of legal frameworks governing weapons and their transfer, with the ability to craft policy that is effective across jurisdictions.

• Problem-Solving in Ambiguous Environments: Comfort with ambiguity and a proactive approach to tackling problems without established playbooks.

• Risk Mitigation Mindset: Motivation to prevent misuse of AI without unduly obstructing legitimate research and engineering in the field.

Preferred Skills:

• Machine Learning & LLM Fundamentals: A working understanding of AI, machine learning principles, and large language models.

• Hands-on Engineering Experience: Practical experience in a weapons-relevant technical domain, such as systems engineering, robotics and autonomy, guidance/navigation/control, sensors and signal processing, aerospace or mechanical engineering, materials and energetics, or embedded software.

• Arms Export Controls Experience: Hands-on experience applying arms export controls (e.g., ITAR, EAR) or international arms-control regimes.

• Classifier Evaluation: Experience building or evaluating classifiers, including LLM-based ones, with an understanding of precision and recall for rare, high-consequence categories.

• Trust & Safety/Product Policy: Prior experience in trust & safety or product policy at a technology platform.

📝 Enhancement Note: The qualifications emphasize a unique blend of domain expertise in conventional weapons and the ability to translate that into AI policy. The "Preferred Skills" highlight a significant advantage for candidates with direct experience in AI/ML, demonstrating the company's commitment to integrating technical AI knowledge with policy creation. The explicit mention of ITAR/EAR and classifier evaluation suggests a need for deep technical and regulatory understanding.

📊 Process & Systems Portfolio Requirements

Portfolio Essentials:

• Policy Framework Design: Showcase examples of developed policy frameworks, demonstrating clarity, operational precision, and adaptability to evolving threats.

• Threat Model & Evaluation Case Studies: Present case studies detailing the methodology used for threat modeling and evaluation of complex systems, highlighting the identification of risks and mitigation strategies.

• Technical Analysis Documentation: Include examples of rigorous technical analysis of complex systems (ideally weapons-related or dual-use) derived from open-source intelligence.

• Cross-Functional Implementation Examples: Demonstrate instances where technical judgments were translated into actionable guidelines or guardrails for engineering or product teams.

• Regulatory Compliance Understanding: Examples illustrating an understanding of relevant legal and regulatory frameworks (e.g., arms control, export regulations) and their application in policy design.

Process Documentation:

• Policy Development Lifecycle: Document the end-to-end process for developing and updating policies, from research and stakeholder engagement to final implementation and monitoring.

• Threat Assessment Methodology: Outline the systematic approach used for identifying, analyzing, and prioritizing potential AI misuse vectors, particularly concerning dual-use technologies.

• Guardrail Implementation Workflow: Detail the process for translating policy requirements into technical specifications for AI model guardrails, detection systems, and enforcement mechanisms.

• Escalation Management Procedures: Illustrate established procedures for handling and resolving complex policy escalations, including communication protocols and decision-making frameworks.

📝 Enhancement Note: For a role focused on policy design and operationalization within AI safety, a portfolio should clearly demonstrate the candidate's ability to think systematically, analyze complex technical domains, and translate abstract principles into concrete, implementable safeguards. The emphasis on "operational precision" and "translation into guardrails" suggests that practical application and system integration are key evaluation criteria.

💵 Compensation & Benefits

Salary Range: $245,000 - $285,000 USD per year.

Benefits:

• Competitive compensation package.

• Optional equity donation matching.

• Generous vacation time.

• Comprehensive parental leave.

• Flexible working hours to support work-life balance.

• Visa sponsorship available for eligible candidates.

Working Hours: 40 hours per week (standard full-time).

📝 Enhancement Note: The salary range is indicative of a senior-level role requiring specialized expertise, particularly in a high-impact and technically complex domain like AI safety and policy. The benefits package is comprehensive, reflecting Anthropic's commitment to employee well-being and professional development, with flexibility being a key component.

🎯 Team & Company Context

🏢 Company Culture

Industry: Artificial Intelligence Research & Development, AI Safety, Public Benefit Corporation.

Company Size: Growing, with an emphasis on "big science" research efforts. This implies a collaborative yet focused environment where individual contributions can have significant impact.

Founded: Anthropic was founded with a mission to create reliable, interpretable, and steerable AI systems, emphasizing safety and societal benefit. This foundational principle likely permeates all aspects of its culture and operations.

Team Structure:

• Safeguards Organization: This role sits within the Safeguards organization, dedicated to building policies, evaluations, and enforcement systems for AI safety.

• Cross-Functional Collaboration: Expect close collaboration with research scientists, AI/ML engineers, software engineers, legal counsel, and product teams. The success of this role hinges on effective communication and integration across these disciplines.

• Research-Driven Environment: The company culture is deeply rooted in empirical AI research, drawing parallels to natural sciences like physics and biology. This suggests a culture that values rigorous analysis, experimentation, and data-driven decision-making.

Methodology:

• Empirical Science Approach: AI research and development are treated as empirical sciences, emphasizing data, experimentation, and iteration.

• Big Science Research: Focus on large-scale, high-impact research efforts rather than fragmented smaller projects.

• Collaborative Research Discussions: Frequent internal discussions to align on and prioritize high-impact work.

• Focus on Impact: Prioritization of work that advances long-term goals of steerable, trustworthy AI.

Company Website: https://www.anthropic.com/

📝 Enhancement Note: Anthropic's culture is defined by its mission-driven approach to AI safety, its commitment to scientific rigor, and its collaborative "big science" ethos. For an operations-focused role like this, it means working within a highly technical and research-oriented environment where policy must be grounded in deep technical understanding and empirical evidence, with a clear focus on real-world impact and risk mitigation.

📈 Career & Growth Analysis

Operations Career Level: This role is positioned at a mid-to-senior level, demanding significant subject matter expertise and the ability to operate with a high degree of autonomy. It sits at the intersection of technical policy, risk management, and operational implementation within a cutting-edge AI company.

Reporting Structure: While not explicitly stated, roles within specialized organizations like "Safeguards" typically report to a Director or Head of Policy/Safeguards, with direct engagement with engineering and product leadership. The role requires significant influence and partnership across teams.

Operations Impact: The Policy Design Manager for Conventional Weapons will have a profound impact on Anthropic's ability to deploy AI responsibly. Their work directly shapes the ethical boundaries and safety guardrails of AI systems, influencing their societal impact and preventing potentially catastrophic misuse. This role is crucial for maintaining public trust and ensuring Anthropic's AI development aligns with its mission of beneficial AI.

Growth Opportunities:

• Specialization Deepening: Opportunity to become a leading expert in AI safety policy, specifically within high-consequence domains like conventional weapons.

• Cross-Functional Leadership: Potential to lead initiatives that bridge policy, engineering, and product development, driving the operationalization of safety measures.

• Mentorship & Team Development: As the Safeguards organization grows, opportunities to mentor junior policy analysts or contribute to building out the team's capabilities.

• Influence on AI Safety Standards: Contribute to shaping industry-wide best practices and standards for AI safety, particularly in sensitive applications.

• Advancement within Safeguards: Potential progression to senior management roles within the Safeguards organization, overseeing broader policy areas or teams.

📝 Enhancement Note: This role offers a unique opportunity for deep specialization in a critical and emerging field of AI safety. Growth is likely to come from deepening expertise, expanding influence across technical and policy domains, and potentially taking on leadership responsibilities within Anthropic's Safeguards organization. The impact is significant, directly contributing to the responsible development and deployment of advanced AI.

🌐 Work Environment

Office Type: Hybrid, with a requirement for in-office presence at least 25% of the time. Offices are located in San Francisco, CA, and New York City, NY, with Washington, DC also being a key location, suggesting a focus on policy and government engagement.

Office Location(s):

• San Francisco, CA, United States

• New York City, NY, United States

• Washington, DC, United States Workspace Context:

• Collaborative Hubs: Offices are designed to facilitate collaboration, likely featuring meeting rooms, project spaces, and common areas conducive to brainstorming and cross-functional work.

• Access to AI Tools & Expertise: While not explicitly detailed, expect access to Anthropic's internal AI tools, research infrastructure, and direct collaboration with leading AI researchers and engineers.

• Policy & Legal Engagement Focus: The presence of a Washington D.C. location suggests an environment that supports engagement with policymakers, regulators, and industry partners in the nation's capital.

Work Schedule: Standard 40-hour work week, with an emphasis on flexible working hours, allowing for adaptation to different project needs and personal schedules, while still maintaining the 25% in-office requirement.

📝 Enhancement Note: The hybrid work model and multiple key locations (SF, NYC, DC) indicate a company that values both in-person collaboration and the flexibility of remote work. The DC presence strongly suggests a focus on the policy, regulatory, and governmental aspects of AI safety, which is critical for a role dealing with conventional weapons policy.

📄 Application & Portfolio Review Process

Interview Process:

• Initial Screening: Review of application, resume, and portfolio by a recruiter and hiring manager to assess alignment with minimum qualifications and preferred skills.

• Technical & Policy Deep Dive: Interviews with members of the Safeguards team and potentially engineering leads to assess subject matter expertise in conventional weapons, policy design capabilities, and understanding of AI safety concepts. This stage may involve scenario-based questions and detailed discussions of past work.

• Portfolio Presentation: A dedicated session where candidates present their portfolio, showcasing relevant case studies, policy frameworks, and technical analyses. This is a critical opportunity to demonstrate practical application of skills.

• Cross-Functional Collaboration Assessment: Interviews with stakeholders from engineering, legal, or product teams to evaluate communication skills, ability to influence, and capacity for cross-functional partnership.

• Values & Culture Fit: Discussions focused on assessing alignment with Anthropic's mission, values (safety, collaboration, impact), and ability to thrive in a research-intensive, mission-driven environment.

Portfolio Review Tips:

• Highlight Policy Creation: Showcase examples of policies you've designed, emphasizing clarity, operationalizability, and the rationale behind them.

• Demonstrate Technical Analysis: Present detailed analyses of complex technical systems, ideally using open-source data, and explain your methodology and findings.

• Illustrate Risk Assessment: Include case studies of threat modeling or risk assessments you've conducted, detailing how you identified vulnerabilities and proposed mitigation strategies.

• Showcase Translation to Action: Provide examples of how you've translated technical insights or policy requirements into concrete actions, guardrails, or detection mechanisms for engineering teams.

• Quantify Impact: Where possible, quantify the impact of your work, e.g., reduction in risk, improvement in policy clarity, or successful implementation of safeguards.

• Tailor to Conventional Weapons: Explicitly connect your experience and portfolio items to the specific domain of conventional weapons and dual-use technologies.

Challenge Preparation:

• Scenario-Based Questions: Be prepared for hypothetical scenarios related to defining boundaries for AI use in sensitive areas, responding to emerging misuse, or translating complex technical risks into policy.

• Technical Deep Dive: Brush up on concepts related to conventional weapons systems, software/autonomy components, and AI safety principles.

• Policy Rationale: Be ready to articulate the "why" behind your policy decisions and demonstrate a balanced approach to safety and utility.

• Communication Practice: Practice explaining complex technical and policy concepts concisely to both technical and non-technical audiences.

📝 Enhancement Note: The application process strongly emphasizes both deep domain expertise and the ability to operationalize that expertise within an AI context. A strong portfolio that demonstrates practical application, clear communication, and a rigorous analytical approach will be crucial for success. The emphasis on "conventional weapons" means candidates must directly address their experience and knowledge in this specific, high-consequence area.

🛠 Tools & Technology Stack

Primary Tools:

• Policy Development & Documentation Platforms: Tools for drafting, versioning, and collaborating on policy documents (e.g., internal wikis, specialized policy management software, advanced document editors).

• Threat Modeling Frameworks: Methodologies and potentially software for visualizing and analyzing complex threat landscapes (e.g., STRIDE, custom frameworks).

• Data Analysis & Visualization Tools: Software for analyzing open-source intelligence and evaluation results (e.g., Python with libraries like Pandas, NumPy; data visualization tools like Tableau, Looker, or internal dashboards).

• Communication & Collaboration Suites: Standard enterprise tools for team communication, project management, and knowledge sharing (e.g., Slack, Google Workspace, Jira).

Analytics & Reporting:

• Evaluation Metrics Tools: Systems for tracking and reporting on the performance of AI model guardrails and detection systems, focusing on precision, recall, and false positive/negative rates for high-consequence categories.

• Reporting Dashboards: Tools for creating and presenting performance metrics to stakeholders, potentially including custom-built dashboards or BI platforms.

CRM & Automation:

• Internal Knowledge Management Systems: Systems for organizing and accessing research, policy documents, and past analyses.

• Workflow Automation (Potential): While not explicitly stated, experience with or understanding of how workflow automation can support policy enforcement or detection processes would be beneficial.

📝 Enhancement Note: While specific tools are not listed, the role implies heavy reliance on analytical tools, documentation platforms, and collaborative software. The core requirement is the ability to apply analytical methodologies and translate findings into policy and actionable safeguards, rather than proficiency in a specific, named software suite beyond standard office and communication tools. Experience with tools for analyzing technical systems and evaluating classifiers (including LLM-based ones) would be highly advantageous.

👥 Team Culture & Values

Operations Values:

• Safety First: A paramount commitment to ensuring AI systems are safe, reliable, and beneficial, with a particular focus on preventing catastrophic misuse.

• Rigorous Analysis: A culture that values deep technical understanding, empirical evidence, and data-driven decision-making in policy formulation and evaluation.

• Impact Orientation: Prioritizing work that contributes significantly to Anthropic's long-term mission of developing trustworthy AI.

• Collaboration & Openness: Encouraging open communication, knowledge sharing, and collaborative problem-solving across diverse teams and disciplines.

• Mission Alignment: A shared dedication to the company's mission of creating beneficial AI systems for society.

Collaboration Style:

• Cross-Functional Partnership: A strong emphasis on working closely with engineering, research, legal, and product teams to ensure policies are technically sound, legally compliant, and practically implementable.

• Evidence-Based Discussions: Debates and decision-making are grounded in technical evidence, rigorous analysis, and a shared understanding of potential risks and benefits.

• Proactive Communication: Encouraging proactive sharing of insights, concerns, and progress across teams to maintain alignment and address challenges swiftly.

• Respect for Expertise: Valuing the diverse expertise of team members, from deep technical knowledge of AI and weapons systems to legal acumen and policy design skills.

📝 Enhancement Note: Anthropic's culture is deeply intertwined with its mission of AI safety. For this role, it means operating in an environment where ethical considerations and risk mitigation are central to all technical and policy decisions. Collaboration is key, requiring individuals who can bridge technical domains and effectively communicate complex ideas to ensure safety measures are robust and well-understood across the organization.

⚡ Challenges & Growth Opportunities

Challenges:

• Defining Dual-Use Boundaries: The inherent difficulty in distinguishing between legitimate civilian/research applications and prohibited weapons development for dual-use technologies presents a significant policy design challenge.

• Rapidly Evolving Technology: Keeping pace with advancements in both AI capabilities and conventional weapons technology requires continuous learning and adaptation of policies and evaluations.

• Translating Technical Judgment to Policy: Effectively converting complex technical evidence and nuanced judgments about weapons systems into clear, actionable, and enforceable policy guidelines.

• Global Regulatory Landscape: Navigating diverse and evolving international legal frameworks and arms control regimes.

• Balancing Safety and Utility: Ensuring robust safeguards are in place without unduly hindering legitimate research, innovation, or applications that could be beneficial.

Learning & Development Opportunities:

• Deepening AI Safety Expertise: Gaining unparalleled experience in a critical and emerging field of AI risk management.

• Domain Specialization: Becoming a recognized expert in the intersection of AI and conventional weapons policy.

• Cross-Disciplinary Skill Development: Enhancing skills in technical analysis, policy design, legal frameworks, and stakeholder management.

• Industry Influence: Opportunity to contribute to shaping AI safety standards and best practices within the broader AI industry and policy community.

• Mentorship: Learning from and potentially mentoring other professionals within Anthropic's Safeguards organization.

📝 Enhancement Note: The challenges in this role are substantial, reflecting the cutting-edge nature of AI safety and the high-consequence domain of conventional weapons. Success will require a proactive, learning-oriented mindset, coupled with the ability to navigate complex technical, ethical, and geopolitical landscapes. The growth opportunities are equally significant, offering a chance to pioneer in a vital area of AI development.

💡 Interview Preparation

Strategy Questions:

• "Describe a time you had to define clear policy boundaries for a complex, dual-use technology. What was your process, and what were the key challenges?" (Focus on your methodology, stakeholder engagement, and how you operationalized the policy.)

• "How would you approach building a threat model for AI's potential contribution to conventional weapons development, given current technological trends?" (Demonstrate your understanding of threat modeling frameworks and your ability to apply them to AI and specific weapon systems.)

• "Imagine you discover a new AI capability that could significantly aid in weapons development. How would you assess the risk, and what steps would you take to translate that assessment into policy and potential guardrails?" (Showcase your risk assessment process, policy formulation skills, and collaboration with engineering.)

• "Explain a complex technical concept related to weapons systems or AI safety to a non-technical audience. What are the key points you would emphasize?" (Practice clear, concise communication, focusing on impact and relevance.) Company & Culture Questions:

• "What motivates you to work on AI safety, specifically in a high-consequence domain like conventional weapons?" (Connect your personal values and career aspirations to Anthropic's mission.)

• "How do you approach balancing the need for safety with the desire for innovation and progress in AI?" (Demonstrate an understanding of Anthropic's nuanced approach to AI development.)

• "Describe your experience working in a highly collaborative, research-driven environment. How do you contribute to team success?" (Highlight your teamwork and communication skills.)

• "Based on your understanding of Anthropic, what do you see as the biggest challenges and opportunities for AI safety in the next five years?" (Show your research and forward-thinking perspective.) Portfolio Presentation Strategy:

• Structure Your Narrative: Organize your portfolio around key themes: policy development, technical analysis, risk assessment, and operationalization.

• Focus on Impact: For each project, clearly articulate the problem, your approach, your specific contributions, and the tangible outcomes or impact achieved. Use metrics where possible.

• Highlight Conventional Weapons Relevance: Explicitly draw connections between your past work and the requirements of this role, even if your previous domain was different. Emphasize transferable skills in complex technical analysis and policy design.

• Demonstrate Process: Walk through your methodologies for threat modeling, policy creation, and technical analysis. Show your thought process.

• Be Ready for Deep Dives: Anticipate detailed questions about your projects, particularly concerning your technical understanding, analytical rigor, and decision-making rationale.

📝 Enhancement Note: Interview preparation should focus on demonstrating a deep understanding of both conventional weapons systems and AI safety principles, coupled with a proven ability to translate these into practical, operational policies. Candidates should be prepared to articulate their decision-making processes and showcase their capacity for rigorous, data-driven analysis in a high-stakes environment.

📌 Application Steps

To apply for this Policy Design Manager, Conventional Weapons position:

• Submit your application through the provided link on Greenhouse.

• Tailor Your Resume: Highlight specific experience in policy design, technical analysis of complex systems (especially weapons or dual-use technologies), risk assessment, and cross-functional collaboration. Use keywords from the job description.

• Curate Your Portfolio: Select 2-3 of your most relevant projects that best demonstrate your capabilities in policy creation, technical analysis (open-source is a plus), and translating technical judgment into actionable guidelines. Ensure these showcase your understanding of risk and your ability to communicate complex ideas.

• Prepare Your Portfolio Presentation: Practice walking through your selected portfolio items, focusing on the problem, your methodology, your contribution, and the impact. Be ready to answer detailed questions about your technical understanding and policy rationale.

• Research Anthropic: Understand Anthropic's mission, its approach to AI safety, its "big science" philosophy, and its stance on beneficial AI. Familiarize yourself with their published research to grasp their technical and ethical direction.

⚠️ Important Notice: This enhanced job description includes AI-generated insights and operations industry-standard assumptions. All details should be verified directly with the hiring organization before making application decisions.