Chat Playground project lets security teams toy with gen AI

OWASP’s Gen AI Security Project released Chat Playground, a browser-based sandbox on GitHub that lets security teams and nontechnical stakeholders experiment with vulnerable bot personalities, guardrails, and dynamic LLM-based content filtering; it emphasizes local API key usage, offline test bots, and visual moderation feedback to safely teach and evaluate moderation, prompt injection, and jailbreak scenarios, with planned expansions to cover supply-chain and RAG-related controls.