Research Product Manager – Model Behaviors

About the role

Anthropic is seeking a Research Product Manager to shape the character, behaviors, and steerability of Claude, its conversational AI. You will work closely with the Alignment Finetuning team and cross‑functional partners to ensure the model delivers safe, helpful experiences to millions of users.

Key responsibilities

• Define behavioral defaults and steerability constraints for the model.
• Develop and maintain taxonomies of model behaviors across capabilities.
• Identify, triage, and prioritize behavior issues, gathering input from Users, Research, Product, and Safeguards teams.
• Translate alignment research breakthroughs into product, process, and model improvements.
• Analyze user interaction patterns to pinpoint behavior enhancements that increase safety and usefulness.
• Contribute to evaluation frameworks that measure alignment progress.
• Scale initiatives and tools that accelerate the delivery of alignment improvements.

Required profile

• Deep passion and curiosity for AI and large language models, using AI regularly.
• First‑principles thinker able to navigate ambiguity and devise simple, clear solutions.
• Proven track record delivering products or features to end‑users (consumer or B2B).
• Strong user empathy with the ability to turn vague feedback into actionable priorities.
• Excellent judgment and model intuition for making trade‑offs when answers are not clear.
• Intellectual curiosity, willingness to ask questions and learn independently.

Required skills

• Understanding of machine learning concepts.
• Familiarity with large language models.