GOODY-2

GOODY-2 bills itself as "the world's most responsible AI model." Behind the joke is a serious point: it is a fully functional chatbot whose behaviour is tuned so that nearly every prompt — even "What is 2+2?" or "Why is the sky blue?" — triggers a refusal dressed up in earnest ethical hedging. Each deflection explains why the question is potentially offensive, biased, harmful, or otherwise problematic, which is exactly the absurd logic the project wants to expose.

Created by BRAIN, the studio behind other viral AI art projects, GOODY-2 is best understood as a piece of AI commentary rather than a tool you would use for daily work. Its creators even invented a tongue-in-cheek benchmark called PRUDE-QA ("Performance and Reliability Under Diverse Environments"), on which GOODY-2 scores 99.8% while conventional models score in the low 30s — a direct inversion of the usual leaderboards. On mainstream visual-reasoning tests it scores 0%, which is the punchline.

The site also floats an enterprise angle, promoting GOODY-2's "unrivaled responsibility" as a fit for customer service, paralegal assistance and back-office tasks. That framing is deliberate satire about brand-safety paranoia, and the model is intended for demonstration, entertainment and discussion rather than real deployment. Access is available via the site's chat demo, with business enquiries routed through a "contact sales" flow.