Turing
AI Safety and Policy Evaluator
Global
TuringPosted Sep 26, 2026
Work Type
Remote
Location
Global
Compensation
Rate on request
Role overview :
We are seeking Subject Matter Experts (SMEs) with expertise in psychology, social work, counseling, qualitative research, HCI, communication, conflict resolution, coaching, or related advisory fields to support AI conversation evaluation research.
SMEs will review conversations between participants and AI models, assess whether sufficient context was gathered before recommendations were made, identify missed opportunities for clarification, and provide structured feedback to improve conversational quality.
Job Responsibilities:
Review AI conversations involving real-world interpersonal and everyday decision-making scenarios.
Identify missing context and recommend relevant follow-up questions.
Evaluate whether additional information would have changed the AI's recommendations.
Distinguish between critical ("load-bearing") and non-essential information.
Conduct qualitative reviews and document findings using structured evaluation guidelines.
Collaborate with research teams to ensure consistent, high-quality assessments..
Required Qualification:
Candidates should have at least 3–5 years of direct practicerelevant qualifications and professional experience in at least one of the following areas:
Licensed Social Work
MSW or equivalent graduate-level qualification in Social Work.
LCSW, LICSW, LMSW, or equivalent applicable social work license/credential.
Experience with client assessment, case review, interviewing, counseling, or related social-work practice.
Counseling, Therapy, or Psychology
MA/MS with LMFT, LPC, LCPC, or equivalent counseling/therapy license or credential; or
PhD/PsyD in Clinical, Counseling, Social, or Industrial-Organizational Psychology.
Experience with counseling, therapy, psychological assessment, interviewing, case review, or other relevant client-facing work.
Qualitative Research, HCI, or Communication
MA/PhD or equivalent advanced expertise in Human-Computer Interaction (HCI), Communication, Anthropology, Sociology, Qualitative Research, or a closely related discipline.
Experience conducting qualitative research, interviews, user research, conversation analysis, thematic analysis, qualitative coding, or similar research activities.
Relevant backgrounds may include Qualitative Research, HCI Research, UX/User Research, Communication Research, Anthropology, Sociology, or related disciplines.
Mediation, Coaching, Student Services, Organizational Consulting, or Related Advisory Fields
Relevant professional qualification, credential, or demonstrated subject-matter expertise.
Significant direct experience in advisory, client-facing, coaching, mediation, student advising/services, organizational consulting, or comparable work.
Experience gathering context, understanding individual or organizational needs, and providing informed recommendations or guidance.
Must be based in Latam, Southeastern Europe, Asia, Africa
Preferred experience:
Experience eliciting relevant context when clients initially provide incomplete information
Experience working across relationships, work, education, family, planning, wellbeing, and resource constraints, considering biological, psychological, social, spiritual/existential, financial, and practical factors
Qualitative coding or structured case review
Ability to distinguish between context that is load-bearing for the recommendation and that is merely descriptive/non-informative
Experience reviewing interviews, consultations, coaching sessions, or client interactions to identify missed opportunities for clarification or information gathering
Project Details :
Project duration - 3 weeks (short term project)
Project type - Pay-per-task (PPT)
AHT - 75 minutes (approx.)
Engagement - Full time
Estimated project start date - 1st October (estimated)
Turing
Global
Turing
Global
Turing
Global
Turing
Global
Turing
Global
Turing
Global
Turing
US
Turing
US
Turing
Global
Join the newsletter to get curated AI roles delivered straight to your inbox, before everyone else.