Work remotely as a French and English AI Safety LLM Evaluator, reviewing model responses, red-teaming safety boundaries, and creating evaluation data. Earn $24 to $36 per hour while helping improve safer AI systems.
Generative AI & RLHF
100% Remote Hourly · $24–$36/hr
$24–$36/hr
Compensation
Worldwide
Eligibility
Intermediate
Experience
Apr 3, 2026
Posted
Open worldwide
Interested in this role?
Create a free OpenTrain account and apply in minutes.
OpenTrain AI is the hiring and contracting organization for this role and the #1 platform for finding and building careers in AI training and data labeling. OpenTrain connects contributors with opportunities to help shape how modern AI systems learn, reason, and respond.
Creating an OpenTrain account is free, and candidates can build a profile and apply in minutes for remote AI-training work.
Fully remote work available worldwide
Hourly contractor engagement
Part-time schedule of 20 or more hours per week
Pay range of $24 to $36 USD per hour
About AI Safety Evaluation
AI training is the human side of building artificial intelligence. People review, rate, annotate, and improve model outputs so AI systems can follow instructions, communicate clearly, and avoid unsafe behavior.
In this role, your evaluations and red-team examples will help establish safety standards for large language models. You will work on cutting-edge technology while applying careful judgment to complex and potentially sensitive content.
Evaluate and rate generated text
Create reinforcement learning and safety-training data
Probe models for harmful or unsafe behaviors
Document patterns that can improve future model responses
The Role
As an AI Safety LLM Evaluator, you will review AI-generated responses and create safety-focused evaluation content in English and French. The work centers on clear reasoning, consistent policy application, and accurate documentation of model behavior.
You will curate red-team training cases across nuanced and potentially explicit content areas. Your expertise will support stronger labeling and safety standards for leading AI models.
Role focus: Large language model safety evaluation
Languages: Near-native or native French and C1 or higher English
Experience level: Intermediate
Data type: Text
Workload: 20 or more hours per week
Employment type: Contractor and part-time
What You’ll Do
You will assess whether AI outputs align with written safety policies and explain your decisions, especially when cases are ambiguous. The role requires close attention to language, context, intent, and potential harm.
Daily work may involve reviewing explicit, toxic, violent, sexual, or psychologically disturbing material. You will use hands-on LLM red-teaming experience to probe safety boundaries and record adversarial patterns.
Score and annotate AI-generated responses
Evaluate outputs for safety, policy alignment, and reasoning quality
Generate safety-focused evaluation content in French and English
Curate red-team cases involving nuanced or potentially explicit content
Identify and document adversarial model behaviors
Apply safety categories consistently across challenging examples
Use tools such as Perplexity, Gemini, ChatGPT, or similar AI systems
Safety Areas You’ll Evaluate
The role requires strong practical knowledge of a broad range of safety categories. You should be able to recognize relevant risks and apply written guidance consistently across different prompts and model responses.
Hate and harassment
Sexual content
Suicide and self-harm
Violence
Bias
Illegal goods and services
Malicious activities
Malicious code
Misinformation
Requirements
Applicants should bring demonstrated experience in trust and safety or a closely related field, together with direct experience testing and evaluating large language models. A bachelor’s degree or higher is expected in a relevant field, although equivalent professional experience may also qualify.
You must be comfortable working in both French and English and reviewing disturbing material as part of your regular responsibilities.
Near-native or native French reading and writing proficiency
Minimum C1 English reading and writing proficiency
Bachelor’s degree or higher in Communications, Linguistics, Psychology, Law or Policy, Security Studies, or a related field, or equivalent professional experience
Proven experience in Trust & Safety, content moderation, policy enforcement, risk operations, investigations, or safety evaluation
Hands-on LLM red-teaming experience, including probing safety boundaries and documenting adversarial patterns
Strong knowledge of the listed AI safety categories
Ability to apply written safety policies consistently
Ability to explain decisions clearly in ambiguous cases
Comfort reviewing explicit, toxic, violent, sexual, or psychologically disturbing content
Prior AI data training, annotation, or evaluation experience preferred
Who Should Apply
This opportunity is designed for an experienced safety, policy, moderation, investigations, or AI-evaluation professional who can combine bilingual language judgment with rigorous red-team analysis. It may suit contributors who want flexible, remote work while directly influencing the behavior of advanced AI systems.
You can work independently in a remote contractor setting
You are confident making and defending nuanced safety judgments
You have practical experience testing LLM safety boundaries
You can sustain at least 20 hours of work each week
You are prepared to engage with sensitive content professionally
How to Apply Through OpenTrain
Create a free OpenTrain account, build your profile around your bilingual safety and LLM red-teaming experience, and apply in minutes. OpenTrain is where people start and grow careers in AI training and data labeling, with remote opportunities across the industry.
Apply as a French and English AI Safety LLM Evaluator
Highlight your red-teaming, trust and safety, and evaluation experience
Review the remote contractor opportunity and supported workload
Begin building your career in the fast-growing AI-training industry
Review and improve French-language AI responses as a remote, hourly contractor. Use expert judgment in French linguistics, reasoning quality, editing, and bilingual communication while earning $16 to $26.50 per hour.
Evaluate and red-team large language models in Hebrew and English, documenting safety failures and policy gaps. This fully remote contractor role pays $26-$38 per hour for 20+ hours weekly.
Use native-level French and careful judgment to evaluate sensitive AI prompts, conversations, and safety behavior. This remote contractor role offers flexible work under 20 hours per week at $48 to $52 per hour.