Это высокоуровневая роль в ведущей AI-лаборатории, предлагающая четкую миссию и значительную ответственность в формировании будущего безопасности и политики в сфере ИИ.
Выше рынка
Трендовый домен
Четкие обязанности
Визовая поддержка
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Подключить Plus
Описание вакансии
Текст:
Hirify AI/Оригинал
TL;DR
Technical Program Manager, Pacing (AI): Running Anthropic's cross-functional Pacing program to coordinate AI safety, policy, research, legal, communications, and engineering commitments with an accent on model training, evaluations, compute, and responsible scaling. Focus on structuring ambiguous programs, driving executive decisions, defining model evidence and sign-off requirements, and maintaining accountability across concurrent projects.
Location: San Francisco, CA or New York City, NY; hybrid policy requires staff to be in an Anthropic office at least 25% of the time
Annual salary: $290,000–$365,000 USD
Company
Anthropic builds reliable, interpretable, and steerable AI systems intended to be safe and beneficial for users and society.
What you will do
Own the Pacing program, maintaining projects, owners, milestones, dependencies, gaps, and leadership visibility.
Turn executive strategy into scoped projects with measurable outcomes, timelines, proposals, and decisions.
Run working sessions, written updates, executive reviews, and follow-up on decisions and action items.
Partner with Responsible Scaling Policy researchers to define model evidence and sign-off requirements across training and deployment.
Coordinate public and internal commitments, publications, and events with Policy and Communications; partner with Legal on confidential work.
Measure progress, recommend changes to programs and practices, and use Claude to build automations.
Requirements
Experience leading multiple complex technical programs across organizations with engineering or research teams, from ambiguous goals through delivery.
Strong program planning, milestone and dependency tracking, and risk management skills across concurrent projects.
Technical fluency to identify risks and challenge tradeoffs in training, evaluating, and deploying frontier AI models.
Working knowledge of AI safety and major debates on developing and governing frontier AI.
Experience delivering through senior leaders and teams without direct reporting authority, including changing how they work.
Experience preparing executive decisions by framing options, recommending an approach, and recording outcomes.
Sound judgment and discretion with confidential and sensitive information.
Bachelor’s degree or equivalent education, training, and/or experience in a relevant field demonstrated through coursework, training, or professional experience.
Nice to have
Experience at a frontier AI developer, government AI institute, independent evaluator, or AI safety or governance research organization.
Hands-on frontier AI safety experience, including safety frameworks, safety cases, risk reports, model evaluations, or training and deployment reviews.
Familiarity with measuring AI progress, including compute, capability trends, and using AI to accelerate AI research.
Experience combining technical work with public policy, coordinating experts across organizations, managing published commitments, or delivering high-visibility releases.
Experience building a program from zero and reshaping it as strategy evolved.
Culture & Benefits
Collaborative research environment focused on large-scale efforts and steerable, trustworthy AI.
Flexible working hours, generous vacation and parental leave, competitive compensation, and optional equity donation matching.
Frequent research discussions and an office environment designed for collaboration.
Visa sponsorship is available, subject to role and candidate circumstances.
Anthropic supports diverse perspectives and encourages qualified candidates to apply even when they do not meet every qualification.
Будьте осторожны
Если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого. Это мошенники. Обязательно жмите «Пожаловаться» или пишите в поддержку. Подробнее в гайде →