All activity
ModernGuard is a fast prompt attack detection API, a fine-tuned modernBERT model. It recognize the most recent, common (and uncommon) prompt injection and jailbreaking patterns ā and continuously learns from your data and feedback.
ModernGuard: LLM guardrail APIFast and adaptive security guardrail for AI agents
sndrnleft a comment
It's a modest launch, fresh ModernGuard MVP is out! --- Why: After working on Apple Siri and building internal guardrails, I realized that securing LLM applications ā especially mission-critical ones ā was a challenging task. Companies and developers face growing risks from prompt injections, jailbreaking attempts, and other adversarial attacks that could compromise AI integrity and business...
ModernGuard: LLM guardrail APIFast and adaptive security guardrail for AI agents