Can Foundation Models Moderate Online Content? Evaluating Instruction- vs. Example-Driven

Our experiments reveal that foundation models can substantially outperform Bluesky's deployed moderation system, nearly tripling its $F_1$ score (0.60 vs. It comes from a paper posted to arXiv on 11 September 2026. The growing complexity of content moderation policies presents a critical challenge for their consistent operationalization. While foundation models possess the basic capabilities needed to confront this challenge, whether they can reliably moderate online content remains an unanswered question. In this paper, we systematically compare two competing paradigms for Vision-Language Model (VLM) guidance: an instruction-driven approach where models reason from policy precepts, and an example-driven approach where they generalize from prior precedents. We ground this investigation in ModerationBench, a new benchmark of 4,000 manually annotated, in-the-wild posts from the Bluesky platform. 0.22) on Random Posts in the benchmark, with both instruction- and example-driven paradigms achieving comparable peak effectiveness. Our findings thus chart a path toward reliable and adaptable policy operationalization at scale.

Why it counts

Our experiments reveal that foundation models can substantially outperform Bluesky's deployed moderation system, nearly tripling its $F_1$ score (0.60 vs.

Sources

Primary source: primary source

What is not known

This brief does not claim independent replication. Claims that appear only on X and not in the primary source stay unknown.