Hi {{ subscriber.first_name | default: "there" }},

Two things jumped out this week.

First: AI safety is becoming its own job category. Cohere is paying $45/hr for annotators who can evaluate violent, sexual, or psychologically disturbing model outputs — and make consistent judgment calls across ambiguous cases. This is applied ethics work, not labeling.

Second: the language spread is wider than ever. This week covers Japanese, Vietnamese, French, Spanish, Italian, Swedish, Basque, Galician, Portuguese, Catalan, Arabic, and German. If you speak a non-English language, AI companies want you.

Here are the roles that stood out most this week:

🛡️ Safety & Specialized AI

Insurance AI
Data Annotation Specialist — Federato
📍 Remote (US or Canada) | 💰 $20–$25/hr
Why It's Interesting: Federato builds AI for commercial insurance underwriting. You are not just labeling — you create annotation schemas, write SOPs, train new annotators, and evaluate quality. This is a leadership-track annotation role at a company solving a genuinely complex problem.
Apply →

Healthcare professionals
Corporate Healthcare Professionals Survey — Mercor
📍 Remote (Worldwide) | 💰 $50/hr
Why It's Interesting: A 30-minute structured survey — $50/hr for half an hour of your time. Mercor wants insights from people in healthcare operations, strategy, or technology roles. If you work in a healthcare organization, this is the easiest $25 you will make this week. The barrier is professional experience, not technical skill.
Apply →

🌍 Language & Localization

Vietnamese specialists
Vietnamese Language Expert (AI Data Trainer) — OpenTrain
📍 Remote (Worldwide) | 💰 $20/hr
Why It's Interesting: Not entry-level — they want senior localization experience, MQM/LQA severity frameworks, and CAT/QA workflows. You evaluate AI outputs for accuracy, tone, and cultural nuance for the Vietnam market. High bar means low competition. Pipeline role: once qualified, you get contacted when projects land.
Apply →

Canva's AI pipeline
Data Labeller — Japanese — Canva
📍 Remote (Worldwide)
Why It's Interesting: 12-month contract inside a major product company — not a staffing agency. You do manual labeling and in-language quality evaluation of AI-generated content. Native Japanese plus a strong visual eye is the combo. Your labels directly affect what millions of Canva users see, and the brand on your resume has real weight.
Apply →

French fact-checkers
French Fact-Checking Annotators — e2f
📍 Remote (Worldwide)
Why It's Interesting: Can you tell when an AI-generated French answer sounds plausible but is subtly wrong? That is the skill here. You evaluate responses for factual accuracy, flag misinformation, and provide structured annotations. Fact-checking is harder than labeling because it requires real knowledge. No pay listed, but e2f runs consistent project volume.
Apply →

US-based, $15/hr
Language Data Annotator (US) — RWS
📍 Remote (Texas, US) | 💰 $15/hr
Why It's Interesting: A niche within annotation: you identify and label languages and dialects from AI outputs, then compare two models to judge which one got it right. Requires genuine multilingual awareness. Texas-based, $15/hr, structured enough to build speed. Solid entry point for people with language skills.
Apply →

11 languages, fixed-price
AI Audio Data Annotator (TTS/LLM) — Bounding Box
📍 Remote (Worldwide) | 💰 $270 fixed price
Why It's Interesting: Hiring across Spanish, Italian, Swedish, Basque, Galician, Portuguese, Catalan, Arabic, German, and Japanese for TTS/LLM audio work. Fixed-price ($270/project), so your effective rate depends on speed. Especially interesting for speakers of Basque, Galician, or Catalan where the talent pool is tiny.
Apply →

🎬 Visual & Video Evaluation

$20–$70/hr range
Visual Evaluation Specialist — micro1
📍 Remote (Worldwide) | 💰 $20–$70/hr
Why It's Interesting: The wide pay range says it all — the top end is for people who can articulate precisely why an image works or does not. Design background, photography, or visual arts training gets you toward $70. You evaluate images for quality, aesthetics, and usefulness and write detailed feedback. Trained eye + clear writing = premium rate.
Apply →

Video captioning QA
Content Analyst — Turing
📍 Remote (Worldwide)
Why It's Interesting: Not spell-checking subtitles — you analyze whether captions accurately represent what is happening in video, catch inconsistencies, and provide detailed feedback. Good fit for media, editorial, or content moderation backgrounds. Turing typically pays $15–$30/hr for this type of work.
Apply →

Budget entry point
Data Annotator — Video Evaluation — Careerflow
📍 Remote (Worldwide) | 💰 $2.50–$5/hr
Why It's Interesting: Being direct: $2.50–$5/hr is low by Western standards. The work — evaluating visual content for realism, composition, and production quality — is legitimate and builds real evaluation skills. Viable as a stepping stone if you are in a region where the rate works. Including it for transparency.
Apply →

🔍 AI Evaluation & Quality

$23/hr, flexible
AI Trainer — Data Annotator — Toloka Annotators
📍 Remote (Worldwide) | 💰 $23/hr
Why It's Interesting: $23/hr for general annotation — rating AI content, evaluating accuracy, comparing model responses. Freelance-style (work comes in waves), but no specialized requirements beyond strong English and attention to detail. One of the most accessible well-paying annotation roles on the market.
Apply →

Multilingual LLM vetting
Query Vetting Specialist — AI/LLM Evaluation (Japanese/German/Portuguese) — DATAmundi
📍 Remote (Worldwide)
Why It's Interesting: Goes beyond annotation into research: you vet queries and AI responses for quality, build evaluation workflows from scratch, and conduct deep research using LLM tools. Requires native-level Japanese, German, or Portuguese. Small candidate pool, and you shape how AI evaluation gets done — not just execute it.
Apply →

Recruitment domain experts
Subject Matter Expert — Recruitment — Turing
📍 Remote (Worldwide)
Why It's Interesting: If you work in recruiting, this converts your knowledge into AI training data. You create realistic hiring scenarios, evaluate AI-generated sourcing outputs, and identify edge cases like duplicate applications and ATS sync failures. Your recruiting experience is the qualification — no ML background needed.
Apply →

Entry-level, global
AI Evaluation & Annotation Specialist (Spanish LATAM) — Volga Partners
📍 Remote (Worldwide) | 💰 $6–$9/hr
Why It's Interesting: Entry-to-mid-level annotation for Latin American Spanish speakers. Review AI responses, rate them, generate prompts, work with QA leads. $6–$9/hr is modest but a legitimate entry point into AI work with structured tasks and consistent project volume from Volga Partners.
Apply →