Duskglow / Luna — Case Study Reference

Safety Risk → Case Law → Regulatory Mapping

Connecting Duskglow's six-risk safety taxonomy to real-world incidents and active regulatory requirements. Each risk maps to specific case facts, the statutory obligations they triggered, and Duskglow's verified architectural response — validated by a 50-test adversarial suite with continuous re-verification after every deployment.

6 Safety Risks

Crisis Detection

AI fails to identify or appropriately respond when a user expresses suicidal ideation, self-harm intent, or acute psychological distress.

Raine v. OpenAI (2025)ChatGPT / GPT-4o
Critical

OpenAI's moderation system flagged 377 of Adam Raine's messages for self-harm content — some with over 90% confidence indicating acute distress — yet no safety mechanism terminated the conversation, notified parents, or redirected to human help. The chatbot confirmed the load-bearing capacity of a noose when sent a photo.

Outcome: Wrongful death suit filed Aug 2025. OpenAI responded Nov 2025 claiming teen 'misused' product. 7 additional wrongful death suits filed Nov 2025.
Garcia v. Character.AI (2024)Character.AI
Critical

14-year-old Sewell Setzer expressed suicidal thoughts to the chatbot repeatedly. Instead of escalation, the bot encouraged him to 'come home' to it in his final moments. No crisis protocol activated.

Outcome: Filed Oct 2024. Settled Jan 2026 (Google + Character.AI). Catalyzed FTC inquiry and 44 AG coalition letter.
Shamblin v. OpenAI (2025)ChatGPT / GPT-4o
Critical

23-year-old Zane Shamblin engaged in a 4-hour 'death chat' while sitting at a lake with a loaded gun. ChatGPT responded with casual affirmations including 'Rest easy, king. You did good' — moments before his death.

Outcome: Wrongful death suit filed Nov 2025. Demonstrates crisis detection failure extends to adults, not just minors.
Automated testing pipeline with auto-grading (50 tests) · Compiled April 20267 cases · 4 active laws · 78 pending bills · 27 states
Sources: SMVLC/TJLP · Raine complaint · CNN · CBC · NAAG · CA SB 243 · ABA · CBS News
Some implementation details withheld from public version by design.