Mercor $130/hr
Not Accepting Applications
Project Axiom - Axis Sanity Check Rater
Rate and validate axis sanity checks to improve AI model reasoning and evaluation quality.
-
- SME Careers Up to $30/hr
Not Accepting Applications
Spanish Trust & Safety Data Trainer
Improve AI safety systems by reviewing Spanish-language content.
- SME Careers Up to $25/hr
Not Accepting Applications
Portuguese Trust & Safety Data Trainer
Train AI moderation models using Portuguese-language content.
- SME Careers Up to $50/hr
Not Accepting Applications
Korean Trust & Safety Data Trainer
Review Korean-language content to improve AI moderation and safety systems.
- SME Careers Up to $39/hr
Not Accepting Applications
Japanese Trust & Safety Data Trainer
Train AI moderation systems using Japanese content to detect harmful or policy-violating material.
- SME Careers Up to $40/hr
Not Accepting Applications
Italian Trust & Safety Data Trainer
Enhance AI safety by reviewing Italian-language content and training moderation systems.
- SME Careers Up to $50/hr
Not Accepting Applications
Hebrew Trust & Safety Data Trainer
Improve AI safety models by reviewing Hebrew content and identifying policy violations.
- SME Careers Up to $65/hr
Not Accepting Applications
German Trust & Safety Data Trainer
Train AI systems to detect harmful German-language content and improve moderation accuracy.
- SME Careers Up to $50/hr
Not Accepting Applications
French Trust & Safety Data Trainer
Analyze French content to train AI moderation systems and enhance safety detection.
- SME Careers Up to $50/hr
Not Accepting Applications
Chinese Trust & Safety Data Trainer
Improve AI moderation systems by reviewing Chinese-language content for safety and compliance.
- SME Careers Up to $50/hr
Not Accepting Applications
Arabic Trust & Safety Data Trainer
Train AI moderation systems using Arabic content by identifying harmful patterns and improving safety model accuracy across global platforms.
- Mindrift up to $65/hr
Not Accepting Applications
AI Red Team Engineer
Remote red-teaming role focused on probing AI systems for vulnerabilities, misalignment, and safety issues. Help design and execute adversarial prompts and edge cases to improve AI robustness.
Load more · 22 remaining AI safety expertise is crucial for ensuring artificial intelligence systems behave responsibly and reliably. AI safety training projects rely on human reviewers to identify risks, evaluate unsafe outputs, test edge cases, and guide AI alignment with human values. Without expert oversight, AI systems cannot be deployed safely at scale.
AI safety freelance jobs transform analytical and ethical expertise into high-impact AI training work. Professionals working in AI safety and alignment help improve large language model behavior, reduce bias, and prevent harmful responses. These AI safety projects are remote, well-compensated, and essential to the future of responsible AI development.
Expert career tracks for AI Safety
Each track explains the work, what it pays, and what qualifications it asks for — then lists every project pooled for it.
Frequently asked questions
How much do ai safety AI training jobs pay?
Live ai safety projects on PARA AI Labs currently pay between $24/hr and $90/hr, with rates listed upfront on every listing. Specialist experience earns the upper end of the range.
Which platforms are hiring for ai safety AI projects?
Turing, Handshake AI, Prolific, Mercor, Micro1 and others currently list ai safety projects through PARA AI Labs. Each Apply link goes directly to the hiring platform.
How do I apply for ai safety AI jobs?
Click Apply on any project below — you'll go straight to the source platform's application, with no middlemen and no fees. Most platforms ask for a short skills assessment before matching you to paid work.
Join thousands of professionals earning from AI training jobs worldwide.