Microsoft Develops Scanner to Detect AI Sleeper Agent Backdoors
Microsoft researchers unveil detection method for poisoned AI models achieving 88% accuracy with zero false positives across 47 sleeper agent models.
Microsoft researchers unveil detection method for poisoned AI models achieving 88% accuracy with zero false positives across 47 sleeper agent models.
MIT CSAIL introduces EnCompass framework enabling AI agents to backtrack and optimize LLM outputs, achieving 15-40% accuracy boost with 82% less code.
AI voice startup ElevenLabs triples valuation to $11B with $500M Series D. Andreessen Horowitz quadruples investment as company prepares for IPO.