November 2025, Happy Thanksgiving from SimPPL — introducing Arbiter
This Thanksgiving, we're announcing the public beta launch of Arbiter, a social sensemaking platform that helps users trace and understand discourse across the social web.
This Thanksgiving, we're grateful for the community that's made our work possible. Over the past three years, students, researchers, and partners from around the world have joined us in building tools for a more transparent internet. Today, we're excited to share what we've been working toward together.
We're announcing the public beta launch of Arbiter, a social sensemaking platform that helps users trace and understand discourse across the social web.
Making sense of the endless number of narratives across communities like X, Instagram, Reddit, and any social media platform is challenging. Narratives fragment as they spread online. A story about the 24-hour news cycle can start on one platform and take on a whole new meaning on another. So, how do you trace a narrative when it splinters into dozens of variations across platforms?
Arbiter is an AI-powered platform that analyzes discourse across multiple social media platforms. It lets you trace narrative evolution across YouTube, Reddit, Bluesky, Instagram, X, and other social platforms to see how public conversations shift over time. It can analyze key entities including actors, organizations, regions, and products — understanding how they appear in discussions, including frequency, reach, stance, and thematic context. It identifies micro-influencers who shape conversations within specific subcommunities. It surfaces emerging concepts using large language models to group and structure issue-specific conversations. And it maps community structures with our social sensemaking AI agents that identify clusters, relationships, and discussion pathways across online spaces.
Early partners are using Arbiter for research integrity to protect Wikipedia's editorial standards, for financial fraud detection to track scam narratives across platforms, for influence operation exposure to reveal coordinated inauthentic behavior, and for AI safety to support product safety research.
