BS Leaderboard
The finest corporate bullshit, ranked by score and displayed for public record. Every entry here was submitted voluntarily by a user who opted in to the leaderboard — nothing appears without explicit consent. The scores are real. The excerpts are real. The authors remain anonymous, which is probably for the best.
Scoring is performed by Claude, Anthropic's AI model, at temperature zero — meaning the same text submitted twice always returns the same score. Each submission is evaluated across 10 categories of corporate non-communication: Synergy Fluff, Fake Urgency, Responsibility Fog, Metric Abuse, Startup Dialect, Virtue Signalling, Padding & Filler, Confidence–Evidence Gap, AI Writing Fingerprint, and Euphemism Body Count. Category scores run from 0 to 100. The overall BSI — Bullshit Index — is the weighted aggregate, expressed as a percentage and assigned a letter grade from A to F.
The scoring system is calibrated against 10 anchor texts — fixed reference submissions with established correct scores that appear in every analysis request. This gives the AI a consistent internal ruler and prevents score drift over time. A press release that scored 87% last week will score 87% next week. Consistency is the point. The leaderboard reflects genuine comparative rankings, not the noise of an uncalibrated system.
The Weekly Top 10 refreshes every seven days. The Hall of Shame is permanent — an all-time record of the highest-scoring submissions ever opted in to the leaderboard. A score above 80% qualifies as Weapons-Grade Bullshit. The current record stands as a testament to what professional communication can achieve when it fully commits to achieving nothing.