Remote AI training and data labeling jobs

Write a Python simulation and a spec sheet with pass/fail thresholds; a frontier model probes your simulation a limited number of times, then submits a design that an agentic grader scores. For control, analog or RF circuit, power electronics or mechanical design experts with a PhD or equivalent industry record. Remote hourly contract, $60–90/hour.

Remote hourly contract for lawyers, policy staff, consultants and technical writers who author long documents in Word on their own Mac. $60–70/hour, paid weekly via Stripe or Wise. Requires your own Microsoft licence and a screen above 2.5 megapixels. Tasks are not described.

Quantum and computational chemistry PhDs write original, runnable research problems for a scientific-computing AI benchmark, with grading criteria, calibrated until frontier models fail them more often than they pass. 6 weeks at 20+ hours a week, Git and Docker workflow. Flat $70/hour; 182 hired this month.

Fuel-cycle and criticality safety engineers red-team frontier AI models: write benign, dual-use and adversarial prompts from enrichment, fabrication, reprocessing or criticality work, judge model replies against a policy standard, and write reference answers. Remote contract at $65–75 per task.

Hands-on preclinical scientists from companies that develop their own drugs annotate R&D data and model outputs, and advise an AI lab building foundation models for drug discovery. Any modality, 5+ years industry R&D, about 10 hours a week. $60–100/hour.

A paid research study for a major tech company building AI design tools: senior product, UX and brand designers take part in recorded remote sessions on how experts judge craft and decide what is ready to ship, either as interviewee or interviewer. About 4–5 hours per session. $150–250/hour, scope still being finalised.

Evaluate AI-generated legal analyses of everyday civil problems (housing, family, consumer, foreclosure, debt) and write structured feedback on reasoning, procedure and access-to-justice issues. For US-admitted lawyers with pro bono program experience. Remote hourly contract at $170/hour.

Remote hourly contract for consultants, analysts and marketers who build decks in PowerPoint on their own Windows PC. $60–70/hour, paid weekly via Stripe or Wise. Your own Microsoft licence and a screen setup above 2.5 megapixels are required; the ad does not describe the tasks.

QA engineers, SDETs and test automation engineers review browser-based test workflows for AI-generated web apps, checking that each test is feasible, isolated, deterministic and gives a clean pass or fail before it joins a benchmark dataset. 3+ years and Playwright, Cypress or Selenium experience. Remote hourly contract at $30–60/hour.

Full-time, on-site-hybrid role in the Bay Area: review AI output on business and sales operations tasks, write instruction specs and golden solutions, and build benchmarks with an AI lab's research team. For senior ops leaders with 4+ years. W-2 via Cincinnatus, $60–100/hour.

Current and former IAEA, Euratom and state-system safeguards inspectors and accountancy analysts red-team frontier AI models: write benign, dual-use and adversarial prompts, judge model replies against a policy standard, and write reference answers. Remote contract at $65–75 per task.

Transcribe professionally recorded Dutch and English audio into clean, accurately punctuated text, proofread it, and annotate linguistic features for AI training. Native Dutch and professional transcription experience expected. Ten openings, contractor, $25–30/hour.

Audit Mandarin Chinese speech data for two Amazon Sonic collections: judge annotators' transcripts against audio using fixed error codes, and fix word-level timestamp alignment. Native Mandarin as spoken in mainland China is a hard requirement; all rules and rationales are in English. Remote hourly contract at $21.50/hour.

Onsite in a San Francisco lab: use a handheld UMI gripper to repeat household manipulation tasks (picking up objects, opening containers) so robots can learn from the recordings. No robotics or AI experience needed. Ten openings, contractor, $10–20/hour.

Audit Korean speech data for Amazon's Sonic project: check transcripts against audio with fixed error codes and pass/fail verdicts, and correct word-level timestamp alignment. Native Korean as spoken in South Korea is a hard requirement; rationales are written in English. Remote hourly contract at $25/hour.

Audit Japanese speech data for Amazon's Sonic collections: verify annotators' transcripts against the audio with fixed error codes and pass/fail calls, and correct word-level timestamp alignment. Native Japanese as spoken in Japan is a hard requirement; rationales are in English. Remote hourly contract at $37.50/hour.

Test AI chat models in Malayalam and English for safety failures and judge whether their answers are accurate, complete and appropriate. Native Malayalam plus careful evaluation skills; prior red-teaming is a plus rather than the core requirement. Remote hourly contract at $16–22/hour, paid weekly.

Full-time W-2 coordinator (via Cincinnatus LLC) placed with a leading AI lab's GenAI team, running day-to-day operations on AI training-data projects: turning leadership input into expert guidelines, onboarding experts, tracking quality and deliverables. US-based, 3+ years of coordination plus hands-on AI data experience. $45–55/hour.

Redline MSAs, NDAs and DPAs against a company playbook, choose the position and the fallback, and write the reasoning behind every markup so an AI can learn it. US-based, with a JD, US bar admission and four years of transactional practice preferred. 50 openings at $90–110/hour.

Use Avid Media Composer and broadcast post-production know-how to label, segment and add metadata to video for AI training, and document your editing decisions. Entry to mid-career editors; one opening, contractor, $15–80/hour.

Preclinical scientists with direct antibody-drug conjugate or bispecific antibody experience annotate R&D data and advise an AI lab building drug discovery foundation models. In-house asset developers only, 5+ years industry R&D, about 10 hours a week. $60–100/hour.

Nuclear medicine physicists, radiopharmacy staff and cyclotron RSOs red-team frontier AI models: write benign, dual-use and adversarial prompts from medical isotope work, judge model replies against a policy standard, and write reference answers. Remote contract at $65–75 per task.

Predict when a named sell-side analyst will publish after a catalyst and what the note will say, reasoning only from a fixed evidence cutoff, then grade AI analyses of the same call. For former lead sell-side analysts or very senior associates in the exact sector, typically 8+ years. $150–250/hour, remote.

Help run LLM training projects on browsing capabilities: track project and annotator performance in Google Sheets, review annotator quality and handle contributor communications. Ops or project management background preferred, but ownership matters more. Remote hourly contract at $50–60/hour, no location limit published.