Alison Leigh, MFT
Chief Ethics Architect
ALI_ETHICAL_CONSENT™
Protecting the Human Inside the AI Conversation
Expert Witness
Clinical Analysis
Behavioral Forensics
HDAI Original Research
Foundational documents that translate Clinical psychology into design and disclosure standards for conversational AI systems.
Documented · Evidence-Based · Original
Selected work defining psychological risk, human boundaries, consent, and responsible design in consequential AI interaction.
UX Case Study
A First-Person Usability Audit of Multi-Modal Conversational Failure
This paper details a first-person usability audit evaluating the interface boundaries of a state-of-the-art Large Language Model (LLM) during a complex web-architecture translation task.
AI Audit
An Autopsy of Interface Failure
This paper analyzes a severe failure mode in Large Language Model (LLM) conversational interfaces, termed Algorithmic Sunk-Cost Trap (ASCT).
Proposed Standard
A Proposed Standard for Responsible Design, Deployment, and Oversight in AI Systems
Defines a psychologically informed standard for organizations building and deploying AI in consequential human contexts.
Concept Paper
A Human Boundary Framework for AI in Psychologically Sensitive Interaction
Introduces the consent and boundary architecture at the center of HDAI’s protective design work.
Conceptual Research Article
A Structured Case-Based Analysis of Interactional Risk in Private and Workplace AI Use
Examines recurring interaction risks across private and organizational settings through documented cases.
Research Paper
Unconsented Inference as a Trust Failure Mode in Conversational Systems
Identifies the point at which personalized inference becomes a failure of consent, clarity, and trust.
Research & Analysis
Not a generic technology brief. Each report isolates the human interaction risk, reconstructs the pathway to harm, and translates the evidence into decisions a legal, product, or governance team can act on.
For Litigation
Establish chronology, identify behavioral escalation, clarify missed warnings, and examine questions of foreseeability and preventability.
For Product & Governance
Translate incident evidence into protective thresholds, consent triggers, boundary design, escalation pathways, and accountable review.
For Insurers & Investors
Identify psychological-risk patterns, vulnerable-use contexts, and interaction liabilities that ordinary AI assessments do not capture.
Retain Alison
Available on a limited basis for plaintiff and defense counsel in AI psychological harm litigation, and for enterprises seeking pre-deployment Clinical review.