Lemma Critical Brief · Category archive
AI Decision Integrity
Hallucination litigation, operational harm from misjudgment, ungrounded outputs.
4 Briefs
SynthID Watermark Reverse-Engineering
How a Statistical Attack Strips the Provenance Mark from AI-Generated Content
McKinsey Lilli's Writable System Prompts
The Layer Governing the AI's Behavior Had No Integrity or Provenance
The Robert Williams Wrongful Arrest
When an AI Face-Match Drove a Government Enforcement Action Without Independent Verification
Noroboto Attack
AI Document Review Input-Integrity Forgery via Embedded Lying Fonts