'Self-State Attacks' New Threat: AI Agents Poisoned via Their Memory(aiweekly.co)
3 pointsby adu_onemore4 hours ago1 comment