Remote AI training and data labeling jobs

Act as ground truth for an AI lab teaching models real enterprise sales work: audit workflows, build golden reference trajectories in a mock sales stack, and refine rubrics. For sellers with around 10 years in enterprise sales. US only, $60–90/hour, placed via Cincinnatus.

Senior mechanical engineers from Fortune 500 or major industrial manufacturers write AI evaluation scenarios drawn from real product design, analysis and certification work, with reference outputs and rubrics on an ASME or ISO track. 5+ years required. Remote hourly contract at $70–80/hour; 285 hires this month.

Formulation and synthesis chemists from pyrotechnics or propellant work red-team frontier AI models: write benign, dual-use and adversarial prompts, grade the model replies against a policy standard, and write reference answers. Remote contract at $65–75 per task.

Senior commercial litigators build AI evaluation scenarios covering discovery, motions, trial and arbitration, with reference pleadings, motions and rubrics. US (FRCP, FRE) or international (English CPR, ICC, LCIA) track. Remote hourly contract at $90–100/hour; 5+ years with trial or arbitration experience.

Certified bomb technicians, EOD veterans and bomb squad leaders red-team frontier AI models: write benign, dual-use and adversarial prompts from public-safety practice, judge model responses against a policy standard, and write reference answers. Remote contract at $65–75 per task.

The best-paid role in Mercor's bilingual AI safety series: Norwegian-speaking PhD chemists and biologists write specialised science prompts and grade how AI models handle dual-use questions. Part-time remote at $77–81/hour. Norway preferred, not required; PhD candidates eligible.

Practising US primary care, family medicine or general internal medicine physicians review outpatient notes and judge AI-written documentation against what a PCP would actually chart. Needs C1 or better in one of 25 listed languages. 2+ years post-residency, 10 hours a week, $170–190/hour.

Adversarial testing of AI chat models in Bahasa Indonesia and English: jailbreaks, prompt injection, bias and multi-turn manipulation, recorded as structured red-team data. Native Indonesian plus prior red-teaming or security experience. Remote hourly contract, $17–25/hour, weekly pay.

Experimental scientists create and review training data for a frontier lab's materials science models: inorganic synthesis, superconductors, semiconductors and advanced packaging, characterization (XRD, SEM, TEM) and fabrication. PhD, MS or equivalent hands-on experience. US-based, 10–40 hours a week, $84/hour.

Produce or judge expert finance deliverables for AI: PnL and AP reconciliations, 3-statement and pro-forma models with purchase accounting, covenant analysis and partnership allocations, all from messy source documents. For product control, FP&A, technical accounting and treasury people with 3+ years. $70/hour, remote.

Solve and evaluate complex electrical design problems (power distribution, protection coordination, equipment sizing, load flow) on a short-term expert evaluation project for a leading technology organization. For practising electrical design engineers. Remote, $70–85/hour.

Hands-on ML researchers take on scoped, open-ended empirical problems: training image classifiers and generators from scratch, fine-tuning open-weight LLMs, adversarial robustness, compression under hard budgets, and multilingual pre-training. 3+ years of ML research (PhD counts). Remote hourly contract at $100–120/hour.

Probe AI chat models and agents for safety failures in Vietnamese and English (jailbreaks, prompt injection, bias, multi-turn manipulation) and document each as reproducible data. Native Vietnamese plus prior red-teaming or security experience. Remote hourly contract, $17–25/hour.

Remote hourly contract for power users, IT administrators and desktop support staff who run Windows 11 Pro as their daily OS. $60–70/hour, paid weekly via Stripe or Wise. Home edition users may not qualify, and a single 1080p screen does not meet the display rule. Tasks are not described.

Export control, treaty and proliferation analysts red-team frontier AI models: write benign, dual-use and adversarial prompts from nonproliferation work, judge model responses against a policy standard, and write reference answers. Remote contract at $65–75 per task.

Full-time W-2 role through Cincinnatus LLC, embedded with a leading AI lab: senior drug development scientists write golden solutions, instruction specs and benchmarks for pharma R&D reasoning. Hybrid in the Bay Area, 40 hours a week for an initial 6 months. PhD, PharmD or MD with 4+ years in industry R&D. $75–115/hour.

Nuclear forensics, radiochemistry and detection specialists red-team frontier AI models: write benign, dual-use and adversarial prompts from characterisation and attribution work, judge model replies against a policy standard, and write reference answers. Remote contract at $65–75 per task.

A full-time W-2 role (via Cincinnatus LLC) embedded with a leading AI lab in the Bay Area: review legal model outputs, write instruction specs and golden solutions, and build legal benchmarks. Hybrid, on-site several days a week, 6-month initial term. $60–100/hour; JD, 5+ years' practice, US bar.

MC&A, physical protection and vulnerability assessment specialists red-team frontier AI models: write benign, dual-use and adversarial prompts from nuclear security work, judge model replies against a policy standard, and write reference answers. Remote contract at $65–75 per task.

Run 20-minute structured interviews that vet senior engineers (GPU kernels, security research, ML compilers) for a frontier lab's AI training and evaluation panel, then tier each candidate and write a summary. For experienced recruiters and technical vetters. US, part-time, about 10 hours a week, $50–60/hour.

Native or native-level French speakers anywhere in the world, with a standard, neutral accent and no regional or foreign influence, record one session of about four hours that a Mercor client uses to clone the voice for its internal customer-experience AI agent. $50–100/hour, no location limit. You are licensing your voice.

Mexico-based native speakers with a Mexican Spanish accent record one session of about four hours, which a Mercor client uses to clone the voice for its internal customer-experience AI agent. $50–100/hour, paid in USD, no acting background required; one hire already this month. You are licensing your voice.

UK-based native English speakers with a neutral, standard British accent record one session of about four hours so a Mercor client can clone the voice for its internal customer-experience AI agent. Open to any gender, no acting experience required. $50–100/hour, paid in USD. You are licensing your voice.

Native Swiss German speakers living in Switzerland record one session of about four hours, which a Mercor client uses to clone the voice for its internal customer-experience AI agent. $50–150/hour, tied for the highest ceiling in the family. No acting experience required. You are licensing your voice.