About the job
Since its inception in 2007, Airbnb has transformed the way people travel and connect. What began with two hosts welcoming three guests in a San Francisco home has blossomed into a global community of over 5 million hosts and more than 2 billion guest arrivals in nearly every country. Our platform enables hosts to offer unique stays and experiences, fostering authentic connections between guests and local communities.
The Community You Will Join:
“Our real innovation is not allowing people to book a home; it’s designing a framework to allow millions of people to trust one another. Trust is the real energy source that drives Airbnb…”
- Brian Chesky, Co-Founder & CEO, Airbnb (2019)
At Airbnb, trust forms the bedrock of our vibrant community. The Content Integrity Data Science team plays a pivotal role in nurturing and safeguarding that trust by ensuring accuracy, authenticity, and compliance with our policies across listings, profiles, messages, and other user-generated content.
We collaborate closely with product, engineering, policy, and operations teams to develop sophisticated, AI-driven systems for content understanding, implement effective human-in-the-loop workflows, and proactively mitigate emerging content risks, enabling confident interactions between guests and hosts.
The Impact You Will Have:
This position is critical in enhancing the safety, trust, and quality of real-world user experiences by refining Airbnb's capabilities to comprehend, interpret, and act on content at scale. You will be instrumental in defining how our platform analyzes listings, profiles, messages, and other user-generated content through the creation of next-generation Trust Content Understanding Models.
The ideal candidate is a proactive and skilled “full-stack” Data Scientist with a strong foundation in applied machine learning and a results-oriented mindset. You will lead significant, high-visibility projects, including:
- Enhancing Airbnb’s content integrity by developing Natural Language Processing (NLP) and Large Language Model (LLM)-based frameworks that effectively assess intent, policy adherence, quality, and risk across listings, profiles, and user communications.
- Creating robust models for identifying problematic or misleading content, focusing on text classification, semantic similarity, information extraction, and generative reasoning for policy interpretation and enforcement.
- Designing and optimizing human-in-the-loop machine learning systems for content review, labeling, escalation, and ongoing model enhancement.
- Establishing systems to promote...

