Safeguards Enforcement Analyst, Safety Evaluations

Summary

About AnthropicAnthropic's mission is to create reliable, interpretable, and steerable AI systems. We want AI to be safe and beneficial for our users and for society as a whole. Our team is a quickly growing group of committed researchers, engineers, policy experts, and business leaders working together to build beneficial AI systems.About the RoleAnthropic's Safeguards team is responsible for enforcing our policies, protecting users, and ensuring our platform is not misused. As a Safeguards Enforcement Analyst focused on Safety Evaluations, you'll play a central role in ensuring our models meet safety and policy standards before and after launch. You'll run and monitor evaluations, drive mitigations when issues surface, coordinate the creation of new evals, and help build the processes and documentation that allow the team to scale this work over time.This role requires someone who is detail-oriented, comfortable navigating ambiguity, and capable of coordinating across teams to break new ground and drive work to completion. This work is deeply cross-functional - you'll partner closely with policy experts, Safeguards engineering teams, and many other stakeholders throughout the organization to ensure our evaluations are comprehensive and current, and that findings translate into meaningful improvements to model behavior.ResponsibilitiesSupport model launch readiness by running evaluations, monitoring and interpreting results, and surfacing regressions or unexpected behavior changes to relevant stakeholdersPartner closely with policy and domain experts throughout the evaluation lifecycle - from identifying risks and scoping the right evaluation approach, to coordinating creation of new evals and ensuring existing ones remain current with evolving policies, threat vectors, and model capabilitiesWork with cross-functional stakeholders to help manage evaluation outcomes, including interpreting results and driving mitigations where neededThink strategically about eval quality to build processes and eval paradigms that keep evaluations unsaturated, high-signal, and insightful as models improveBuild out processes and frameworks for creating product-specific evaluations as Anthropic's product surface area expandsHelp design and scope tooling improvements that accommodate evolving eval needs and expand self-serve eval creation and iteration for non-technical usersWrite and maintain rigorous documentation for evaluation creation, execution, and interpretation as the team builds out eval tooling and processesYou may be a good fit if you:Have experience in trust and safety, content operations, policy enforcement, or a related operational role at a technology companyThrive in ambiguous, fast-moving environments - you're energized rather than frustrated when the path forward isn't clearly defined and you need to figure it out as you goHave experience building processes, workflows, or programs from scratch (zero-to-one work), not just maintaining existing onesHave strong program management instincts, naturally creating structure around complex, multi-stakeholder efforts by tracking timelines, dependencies, and deliverables to keep work on trackAre eager to expand your technical toolkit, including adopting internal tools and AI-assisted workflows (e.g., Claude Code) to accelerate your workCan manage multiple concurrent workstreams across different domain areas without losing track of details - strong prioritization and context-switching are essential when deadlines and priorities shift quicklyAre a strong generalist comfortable moving fluidly across different types of work and switching contexts throughout the dayAre comfortable making judgment calls with incomplete information and escalating appropriately when neededCommunicate clearly and concisely, both in writing and cross-functionallyStrong candidates may also have:Experience operating under tight, high-stakes timelines - such as product launch cycles, incident response, or regulatory deadlines - where information and priorities can shift with little noticeExperience coordinating across engineering, policy, and product teams to translate findings into concrete actionExperience building and maintaining SOPs, runbooks, and operational documentation in fast-changing environmentsProficiency with data tools (SQL, dashboards, spreadsheets) sufficient to maintain and improve workflowsComfort working with sensitive content areas as part of eval creation or enforcement review responsibilitiesThe annual compensation range for this role is listed below.For sales roles, the range provided is the role's On Target Earnings ("OTE") range, meaning that the range includes both the sales commissions/sales bonuses target and annual base salary for the role.Annual Salary: $230,000 - $270,000 USDLogisticsMinimum education: Bachelor's degree or an equivalent combination of education, training, and/or experienceRequired field of study: A field relevant to the role as demonstrated through coursework, training, or professional experienceMinimum years of experience: Years of experience required will correlate with the internal job level requirements for the positionLocation-based hybrid policy: Currently, we expect all staff to be in one of our offices at least 25% of the time. However, some roles may require more time in our offices.Visa sponsorship: We do sponsor visas! However, we aren't able to successfully sponsor visas for every role and every candidate. But if we make you an offer, we will make every reasonable effort to get you a visa, and we retain an immigration lawyer to help with this.How we're differentWe believe that the highest-impact AI research will be big science. At Anthropic we work as a single cohesive team on just a few large-scale research efforts. And we value impact - advancing our long-term goals of steerable, trustworthy AI - rather than work on smaller and more specific puzzles. We view AI research as an empirical science, which has as much in common with physics and biology as with traditional efforts in computer science. We're an extremely collaborative group, and we host frequent research discussions to ensure that we are pursuing the highest-impact work at any given time. As such, we greatly value communication skills.The easiest way to understand our research directions is to read our recent research. This research continues many of the directions our team worked on prior to Anthropic, including: GPT-3, Circuit-Based Interpretability, Multimodal Neurons, Scaling Laws, AI & Compute, Concrete Problems in AI Safety, and Learning from Human Preferences.Come work with us!Anthropic is a public benefit corporation headquartered in San Francisco. Guidance on Candidates' AI Usage: Learn about our policy for using AI in our application process.competitive compensation and benefitsoptional equity donation matchinggenerous vacation and parental leaveflexible working hourslovely office space in which to collaborate with colleagues#J-18808-Ljbffr

Job Description

About AnthropicAnthropic's mission is to create reliable, interpretable, and steerable AI systems. We want AI to be safe and beneficial for our users and for society as a whole. Our team is a quickly growing group of committed researchers, engineers, policy experts, and business leaders working together to build beneficial AI systems.About the RoleAnthropic's Safeguards team is responsible for enforcing our policies, protecting users, and ensuring our platform is not misused. As a Safeguards Enforcement Analyst focused on Safety Evaluations, you'll play a central role in ensuring our models meet safety and policy standards before and after launch. You'll run and monitor evaluations, drive mitigations when issues surface, coordinate the creation of new evals, and help build the processes and documentation that allow the team to scale this work over time.This role requires someone who is detail-oriented, comfortable navigating ambiguity, and capable of coordinating across teams to break new ground and drive work to completion. This work is deeply cross-functional - you'll partner closely with policy experts, Safeguards engineering teams, and many other stakeholders throughout the organization to ensure our evaluations are comprehensive and current, and that findings translate into meaningful improvements to model behavior.ResponsibilitiesSupport model launch readiness by running evaluations, monitoring and interpreting results, and surfacing regressions or unexpected behavior changes to relevant stakeholdersPartner closely with policy and domain experts throughout the evaluation lifecycle - from identifying risks and scoping the right evaluation approach, to coordinating creation of new evals and ensuring existing ones remain current with evolving policies, threat vectors, and model capabilitiesWork with cross-functional stakeholders to help manage evaluation outcomes, including interpreting results and driving mitigations where neededThink strategically about eval quality to build processes and eval paradigms that keep evaluations unsaturated, high-signal, and insightful as models improveBuild out processes and frameworks for creating product-specific evaluations as Anthropic's product surface area expandsHelp design and scope tooling improvements that accommodate evolving eval needs and expand self-serve eval creation and iteration for non-technical usersWrite and maintain rigorous documentation for evaluation creation, execution, and interpretation as the team builds out eval tooling and processesYou may be a good fit if you:Have experience in trust and safety, content operations, policy enforcement, or a related operational role at a technology companyThrive in ambiguous, fast-moving environments - you're energized rather than frustrated when the path forward isn't clearly defined and you need to figure it out as you goHave experience building processes, workflows, or programs from scratch (zero-to-one work), not just maintaining existing onesHave strong program management instincts, naturally creating structure around complex, multi-stakeholder efforts by tracking timelines, dependencies, and deliverables to keep work on trackAre eager to expand your technical toolkit, including adopting internal tools and AI-assisted workflows (e.g., Claude Code) to accelerate your workCan manage multiple concurrent workstreams across different domain areas without losing track of details - strong prioritization and context-switching are essential when deadlines and priorities shift quicklyAre a strong generalist comfortable moving fluidly across different types of work and switching contexts throughout the dayAre comfortable making judgment calls with incomplete information and escalating appropriately when neededCommunicate clearly and concisely, both in writing and cross-functionallyStrong candidates may also have:Experience operating under tight, high-stakes timelines - such as product launch cycles, incident response, or regulatory deadlines - where information and priorities can shift with little noticeExperience coordinating across engineering, policy, and product teams to translate findings into concrete actionExperience building and maintaining SOPs, runbooks, and operational documentation in fast-changing environmentsProficiency with data tools (SQL, dashboards, spreadsheets) sufficient to maintain and improve workflowsComfort working with sensitive content areas as part of eval creation or enforcement review responsibilitiesThe annual compensation range for this role is listed below.For sales roles, the range provided is the role's On Target Earnings ("OTE") range, meaning that the range includes both the sales commissions/sales bonuses target and annual base salary for the role.Annual Salary: $230,000 - $270,000 USDLogisticsMinimum education: Bachelor's degree or an equivalent combination of education, training, and/or experienceRequired field of study: A field relevant to the role as demonstrated through coursework, training, or professional experienceMinimum years of experience: Years of experience required will correlate with the internal job level requirements for the positionLocation-based hybrid policy: Currently, we expect all staff to be in one of our offices at least 25% of the time. However, some roles may require more time in our offices.Visa sponsorship: We do sponsor visas! However, we aren't able to successfully sponsor visas for every role and every candidate. But if we make you an offer, we will make every reasonable effort to get you a visa, and we retain an immigration lawyer to help with this.How we're differentWe believe that the highest-impact AI research will be big science. At Anthropic we work as a single cohesive team on just a few large-scale research efforts. And we value impact - advancing our long-term goals of steerable, trustworthy AI - rather than work on smaller and more specific puzzles. We view AI research as an empirical science, which has as much in common with physics and biology as with traditional efforts in computer science. We're an extremely collaborative group, and we host frequent research discussions to ensure that we are pursuing the highest-impact work at any given time. As such, we greatly value communication skills.The easiest way to understand our research directions is to read our recent research. This research continues many of the directions our team worked on prior to Anthropic, including: GPT-3, Circuit-Based Interpretability, Multimodal Neurons, Scaling Laws, AI & Compute, Concrete Problems in AI Safety, and Learning from Human Preferences.Come work with us!Anthropic is a public benefit corporation headquartered in San Francisco. Guidance on Candidates' AI Usage: Learn about our policy for using AI in our application process.competitive compensation and benefitsoptional equity donation matchinggenerous vacation and parental leaveflexible working hourslovely office space in which to collaborate with colleagues#J-18808-Ljbffr