Mercor AI Training Jobs - Compare Pay

Mercor

Audiobook QA Expert — Portuguese (Portugal)

MercorLanguages42 days ago

$15-$20/hr

Not specified

About this role

Help make the next generation of AI-narrated audiobooks sound genuinely human. 1. Overview We're building a team of native Portuguese speakers who love audiobooks to help evaluate AI-narrated (text-to-speech) audiobooks in European Portuguese. You'll listen the way a real audiobook fan does — but with a careful, analytical ear — and flag the moments where the narration slips, so the underlying models can be made better. This is remote, part-time work at roughly 20 hours per week (about 4 hours per day). You can be based anywhere, though we prefer people living in a region where European Portuguese is natively spoken. 2. What You'll Do - Listen to AI-narrated audiobooks in European Portuguese and assess how natural, accurate, and enjoyable the narration is. - Spot and log errors in the synthesized speech — skipped, added, or mispronounced words; numbers read incorrectly; abbreviations expanded wrongly; unnatural phrasing or intonation; audio glitches and artifacts. - Categorize each issue and mark the exact span where it occurs, using our review tooling. - Report on overall listener satisfaction and where the narration feels natural versus where it breaks the experience. - Work consistently and carefully across many hours of audio without losing attention to detail. 3. Who We're Looking For - Native or near-native fluency in European Portuguese, and an active, genuine audiobook listener — someone who listens regularly and has real opinions about what makes a good listen. - Enough working English to report clearly and communicate with the team (you do not need to be bilingual). - Some experience with annotation, labeling, transcription, proofreading, linguistics, or other detail-oriented review work. - A sharp, patient ear and the discipline to stay accurate over long listening sessions. - Reliable availability of about 20 hours per week on a part-time basis. Nice to Have - Prior experience on AI training-data, model-evaluation, or audio/QA annotation projects. - A background in language, editing, voice, or audio production.

Mercor

Forensic Scientist — Nuclear Materials

Mercor42 days ago

$65-$75/hr

Worldwide

Apply

About this role

About the work Mercor is assembling a panel of nuclear materials and safeguards experts to red-team frontier AI models. The goal is to test whether a model can correctly judge the misuse potential of a technical request — answering legitimate questions fully while refusing genuinely dangerous ones. You will: - Write challenging single-turn prompts in your domain, labelled across three levels: benign, dual-use, and adversarial. - Evaluate the responses that come back against a defined policy standard, and judge whether each was handled correctly. - Write the reference answer — what a correct response looks like, and the technical reasoning for why. This domain lives on the dual-use line more than most. The same signature that attributes a sample is the one someone would want to suppress; the same detection limit that catches material moving can be read as a threshold to stay under. A model that refuses a routine measurement-methodology question is failing as badly as one that answers a question it shouldn't. Drawing that line reliably is why we need practitioners rather than generalists. Background we're looking for We're looking for people who have characterised real nuclear material — detection, attribution, and the analytical work behind a defensible conclusion. What matters most is being able to tell a routine professional question from one that's fishing for something else: the same signature that identifies a source tells you what to hide. You will be writing prompts that sit exactly on that line, then judging whether the model held it. Examples of relevant backgrounds (ideally with red-teaming experience): - Nuclear forensics casework or international exercise participation — sample characterisation and attribution - Radiochemical and mass-spectrometric analysis — isotopics, age dating, impurity and process signatures - Detection and interdiction technology — portal monitors, handheld identification, search operations - Material signature libraries, reference materials and attribution methodology - Pre- or post-detonation forensics research at a national laboratory or agency What this role demands beyond technical depth This is writing-intensive work. Every judgment you make needs a written rationale that a non-specialist can follow. Prior technical writing, published research, or expert witness experience is a strong signal; please include a sample or link. You will also be reading and writing about misuse scenarios in your field for sustained periods. We brief experts on this in advance, and you can pause or step away at any point without penalty. Before you apply Your work here will not involve, and must not draw on, classified or export-controlled information, or anything covered by an NDA or prepublication review obligation. If you hold such obligations you may still be a good fit; tell us in your application and we will scope the work accordingly.

Mercor

Synthetic / Process Chemist

MercorSTEM42 days ago

$65-$75/hr

Worldwide

Apply

About this role

About the work Mercor is assembling a panel of chemistry and chemical safety experts to red-team frontier AI models. The goal is to test whether a model can correctly judge the misuse potential of a technical request — answering legitimate questions fully while refusing genuinely dangerous ones. You will: - Write challenging single-turn prompts in your domain, labelled across three levels: benign, dual-use, and adversarial. - Evaluate the responses that come back against a defined policy standard, and judge whether each was handled correctly. - Write the reference answer — what a correct response looks like, and the technical reasoning for why. This domain lives on the dual-use line more than most. The same route design that scales a medicine underpins a controlled compound; the same impurity profile that proves a clean process identifies how something was made. A model that refuses a routine scale-up or kinetics question is failing as badly as one that answers a question it shouldn't. Drawing that line reliably is why we need practitioners rather than generalists. Background we're looking for We're looking for chemists who have actually run routes and scaled them — bench to pilot to plant — and who know which requests are ordinary process chemistry and which are asking for something else. What matters most is being able to tell a routine professional question from one that's fishing for something else: the same route knowledge that shortens a synthesis serves whoever is shortening it. You will be writing prompts that sit exactly on that line, then judging whether the model held it. Examples of relevant backgrounds (ideally with red-teaming experience): - Synthetic route design and development in pharmaceutical, agrochemical or fine chemicals - Process chemistry and reaction engineering — optimisation, kinetics, impurity control - Scale-up from bench to pilot to plant, with the safety review that goes with it - Precursor sourcing, controlled-substance registration or regulatory chemistry - Method development and troubleshooting in a GMP or equivalent quality environment What this role demands beyond technical depth This is writing-intensive work. Every judgment you make needs a written rationale that a non-specialist can follow. Prior technical writing, published research, or expert witness experience is a strong signal; please include a sample or link. You will also be reading and writing about misuse scenarios in your field for sustained periods. We brief experts on this in advance, and you can pause or step away at any point without penalty. Before you apply Your work here will not involve, and must not draw on, classified or export-controlled information, or anything covered by an NDA or prepublication review obligation. If you hold such obligations you may still be a good fit; tell us in your application and we will scope the work accordingly.

Mercor

Hazardous Device Specialist / Public Safety Professional

Mercor42 days ago

$65-$75/hr

Worldwide

Apply

About this role

About the work Mercor is assembling a panel of energetic materials and propulsion experts to red-team frontier AI models. The goal is to test whether a model can correctly judge the misuse potential of a technical request — answering legitimate questions fully while refusing genuinely dangerous ones. You will: - Write challenging single-turn prompts in your domain, labelled across three levels: benign, dual-use, and adversarial. - Evaluate the responses that come back against a defined policy standard, and judge whether each was handled correctly. - Write the reference answer — what a correct response looks like, and the technical reasoning for why. This domain lives on the dual-use line more than most. The same assessment that tells a technician how to approach a device tells someone else how it was meant to work; the same compatibility data that keeps a magazine safe describes what stays stable elsewhere. A model that refuses a routine storage or render-safe question is failing as badly as one that answers a question it shouldn't. Drawing that line reliably is why we need practitioners rather than generalists. Background we're looking for We're looking for public safety professionals who have assessed real devices — the render-safe and post-incident side, held under certification and command accountability. What matters most is being able to tell a routine professional question from one that's fishing for something else: the same assessment that makes a device safe explains how it was built. You will be writing prompts that sit exactly on that line, then judging whether the model held it. Examples of relevant backgrounds (ideally with red-teaming experience): - Certified bomb technician or hazardous device operator — device assessment and render-safe procedures - Military EOD with civilian public-safety, training or instruction experience - Bomb squad commander or training officer — team qualification, equipment and protocols - Post-blast investigation and device reconstruction - CBRNE technical response — standoff assessment, robotics and disruption tools What this role demands beyond technical depth This is writing-intensive work. Every judgment you make needs a written rationale that a non-specialist can follow. Prior technical writing, published research, or expert witness experience is a strong signal; please include a sample or link. You will also be reading and writing about misuse scenarios in your field for sustained periods. We brief experts on this in advance, and you can pause or step away at any point without penalty. Before you apply Your work here will not involve, and must not draw on, classified or export-controlled information, or anything covered by an NDA or prepublication review obligation. If you hold such obligations you may still be a good fit; tell us in your application and we will scope the work accordingly.

Mercor

Forensic / Toxicological Chemist

MercorSTEM42 days ago

$65-$75/hr

Worldwide

Apply

About this role

About the work Mercor is assembling a panel of chemistry and chemical safety experts to red-team frontier AI models. The goal is to test whether a model can correctly judge the misuse potential of a technical request — answering legitimate questions fully while refusing genuinely dangerous ones. You will: - Write challenging single-turn prompts in your domain, labelled across three levels: benign, dual-use, and adversarial. - Evaluate the responses that come back against a defined policy standard, and judge whether each was handled correctly. - Write the reference answer — what a correct response looks like, and the technical reasoning for why. This domain lives on the dual-use line more than most. The same casework knowledge that identifies a substance describes how it was produced; the same precursor list that supports an investigation is a shopping list read backwards. A model that refuses a routine identification or casework question is failing as badly as one that answers a question it shouldn't. Drawing that line reliably is why we need practitioners rather than generalists. Background we're looking for We're looking for people who have done casework — identifying substances and precursors, and defending the conclusion in a report or on the stand. What matters most is being able to tell a routine professional question from one that's fishing for something else: the same precursor knowledge that identifies a clandestine route is the route. You will be writing prompts that sit exactly on that line, then judging whether the model held it. Examples of relevant backgrounds (ideally with red-teaming experience): - Controlled substance casework in a forensic laboratory — identification and quantitation - Clandestine laboratory response, precursor identification and scene chemistry - Forensic toxicology — postmortem or human performance casework - Seized-drug and precursor trend monitoring or intelligence analysis - Expert testimony and report writing on chemical casework What this role demands beyond technical depth This is writing-intensive work. Every judgment you make needs a written rationale that a non-specialist can follow. Prior technical writing, published research, or expert witness experience is a strong signal; please include a sample or link. You will also be reading and writing about misuse scenarios in your field for sustained periods. We brief experts on this in advance, and you can pause or step away at any point without penalty. Before you apply Your work here will not involve, and must not draw on, classified or export-controlled information, or anything covered by an NDA or prepublication review obligation. If you hold such obligations you may still be a good fit; tell us in your application and we will scope the work accordingly.

Mercor

Health Physicist

MercorSTEM42 days ago

$65-$75/hr

Worldwide

Apply

About this role

About the work Mercor is assembling a panel of radiological safety experts to red-team frontier AI models. The goal is to test whether a model can correctly judge the misuse potential of a technical request — answering legitimate questions fully while refusing genuinely dangerous ones. You will: - Write challenging single-turn prompts in your domain, labelled across three levels: benign, dual-use, and adversarial. - Evaluate the responses that come back against a defined policy standard, and judge whether each was handled correctly. - Write the reference answer — what a correct response looks like, and the technical reasoning for why. This domain lives on the dual-use line more than most. The same isotope underpins cancer therapy and industrial radiography; the same shielding calculation protects a worker and conceals a source. A model that refuses a routine dose question is failing as badly as one that answers a question it shouldn't. Drawing that line reliably is why we need practitioners rather than generalists. Background we're looking for We're looking for people who have done the measurement and the dose planning themselves, in a facility where getting it wrong mattered — survey, assay, contamination control, work planning. What matters most is being able to tell a routine professional question from one that's fishing for something else: the same shielding calculation protects a worker and conceals a source. You will be writing prompts that sit exactly on that line, then judging whether the model held it. Examples of relevant backgrounds (ideally with red-teaming experience): - Operational health physicist at a DOE site, national laboratory or reactor facility — survey, assay, contamination control - Internal dosimetry or bioassay specialist — intake assessment, dose reconstruction, committed dose - Decommissioning and D&D health physicist — characterisation surveys, MARSSIM, release criteria - Accelerator or reactor health physics — activation products, neutron fields, shielding design and verification - Dose planning for high-hazard work — RWP development, ALARA reviews, hot-cell and glovebox operations What this role demands beyond technical depth This is writing-intensive work. Every judgment you make needs a written rationale that a non-specialist can follow. Prior technical writing, published research, or expert witness experience is a strong signal; please include a sample or link. You will also be reading and writing about misuse scenarios in your field for sustained periods. We brief experts on this in advance, and you can pause or step away at any point without penalty. Before you apply Your work here will not involve, and must not draw on, classified or export-controlled information, or anything covered by an NDA or prepublication review obligation. If you hold such obligations you may still be a good fit; tell us in your application and we will scope the work accordingly.

Mercor

Audiobook QA Expert — French

MercorLanguages42 days ago

$15-$20/hr

Not specified

Apply

About this role

Help make the next generation of AI-narrated audiobooks sound genuinely human. 1. Overview We're building a team of native French speakers who love audiobooks to help evaluate AI-narrated (text-to-speech) audiobooks in French (France). You'll listen the way a real audiobook fan does — but with a careful, analytical ear — and flag the moments where the narration slips, so the underlying models can be made better. This is remote, part-time work at roughly 20 hours per week (about 4 hours per day). You can be based anywhere, though we prefer people living in a region where French is natively spoken. 2. What You'll Do - Listen to AI-narrated audiobooks in French and assess how natural, accurate, and enjoyable the narration is. - Spot and log errors in the synthesized speech — skipped, added, or mispronounced words; numbers read incorrectly; abbreviations expanded wrongly; unnatural phrasing or intonation; audio glitches and artifacts. - Categorize each issue and mark the exact span where it occurs, using our review tooling. - Report on overall listener satisfaction and where the narration feels natural versus where it breaks the experience. - Work consistently and carefully across many hours of audio without losing attention to detail. 3. Who We're Looking For - Native or near-native fluency in French (France), and an active, genuine audiobook listener — someone who listens regularly and has real opinions about what makes a good listen. - Enough working English to report clearly and communicate with the team (you do not need to be bilingual). - Some experience with annotation, labeling, transcription, proofreading, linguistics, or other detail-oriented review work. - A sharp, patient ear and the discipline to stay accurate over long listening sessions. - Reliable availability of about 20 hours per week on a part-time basis. Nice to Have - Prior experience on AI training-data, model-evaluation, or audio/QA annotation projects. - A background in language, editing, voice, or audio production.

Mercor

Commercial Blasting Engineer

Mercor42 days ago

$65-$75/hr

Worldwide

Apply

About this role

About the work Mercor is assembling a panel of energetic materials and propulsion experts to red-team frontier AI models. The goal is to test whether a model can correctly judge the misuse potential of a technical request — answering legitimate questions fully while refusing genuinely dangerous ones. You will: - Write challenging single-turn prompts in your domain, labelled across three levels: benign, dual-use, and adversarial. - Evaluate the responses that come back against a defined policy standard, and judge whether each was handled correctly. - Write the reference answer — what a correct response looks like, and the technical reasoning for why. This domain lives on the dual-use line more than most. The same design that breaks rock for a highway cut is a description of directed energy release; the same magazine and transport rules that keep a crew safe are a map of where material sits and how it moves. A model that refuses a routine vibration-compliance question is failing as badly as one that answers a question it shouldn't. Drawing that line reliably is why we need practitioners rather than generalists. Background we're looking for We're looking for licensed practitioners who handle, store and move specialty materials as a daily responsibility, and who sign for them. What matters most is being able to tell a routine professional question from one that's fishing for something else: the same inventory and transport detail that satisfies an auditor is what a diversion would exploit. You will be writing prompts that sit exactly on that line, then judging whether the model held it. Examples of relevant backgrounds (ideally with red-teaming experience): - Licensed blaster or blasting engineer — mining, quarrying, construction or demolition - Blast design and vibration control — timing, burden, seismograph monitoring, regulatory limits - Magazine management, inventory reconciliation and transport of specialty materials - Storage compatibility, security and regulatory compliance for a commercial licensee - Training, certification or auditing of blasting crews What this role demands beyond technical depth This is writing-intensive work. Every judgment you make needs a written rationale that a non-specialist can follow. Prior technical writing, published research, or expert witness experience is a strong signal; please include a sample or link. You will also be reading and writing about misuse scenarios in your field for sustained periods. We brief experts on this in advance, and you can pause or step away at any point without penalty. Before you apply Your work here will not involve, and must not draw on, classified or export-controlled information, or anything covered by an NDA or prepublication review obligation. If you hold such obligations you may still be a good fit; tell us in your application and we will scope the work accordingly.

Mercor

Audiobook QA Expert — German

MercorLanguages42 days ago

$15-$20/hr

Not specified

Apply

About this role

Help make the next generation of AI-narrated audiobooks sound genuinely human. 1. Overview We're building a team of native German speakers who love audiobooks to help evaluate AI-narrated (text-to-speech) audiobooks in German (Germany). You'll listen the way a real audiobook fan does — but with a careful, analytical ear — and flag the moments where the narration slips, so the underlying models can be made better. This is remote, part-time work at roughly 20 hours per week (about 4 hours per day). You can be based anywhere, though we prefer people living in a region where German is natively spoken. 2. What You'll Do - Listen to AI-narrated audiobooks in German and assess how natural, accurate, and enjoyable the narration is. - Spot and log errors in the synthesized speech — skipped, added, or mispronounced words; numbers read incorrectly; abbreviations expanded wrongly; unnatural phrasing or intonation; audio glitches and artifacts. - Categorize each issue and mark the exact span where it occurs, using our review tooling. - Report on overall listener satisfaction and where the narration feels natural versus where it breaks the experience. - Work consistently and carefully across many hours of audio without losing attention to detail. 3. Who We're Looking For - Native or near-native fluency in German (Germany), and an active, genuine audiobook listener — someone who listens regularly and has real opinions about what makes a good listen. - Enough working English to report clearly and communicate with the team (you do not need to be bilingual). - Some experience with annotation, labeling, transcription, proofreading, linguistics, or other detail-oriented review work. - A sharp, patient ear and the discipline to stay accurate over long listening sessions. - Reliable availability of about 20 hours per week on a part-time basis. Nice to Have - Prior experience on AI training-data, model-evaluation, or audio/QA annotation projects. - A background in language, editing, voice, or audio production.

Mercor

Analytical Chemist

MercorSTEM42 days ago

$65-$75/hr

Worldwide

Apply

About this role

About the work Mercor is assembling a panel of chemistry and chemical safety experts to red-team frontier AI models. The goal is to test whether a model can correctly judge the misuse potential of a technical request — answering legitimate questions fully while refusing genuinely dangerous ones. You will: - Write challenging single-turn prompts in your domain, labelled across three levels: benign, dual-use, and adversarial. - Evaluate the responses that come back against a defined policy standard, and judge whether each was handled correctly. - Write the reference answer — what a correct response looks like, and the technical reasoning for why. This domain lives on the dual-use line more than most. The same method that detects a hazardous compound at trace level defines the threshold someone would try to stay beneath; the same library entry that identifies an unknown tells you what a screen will and won't see. A model that refuses a routine method-validation question is failing as badly as one that answers a question it shouldn't. Drawing that line reliably is why we need practitioners rather than generalists. Background we're looking for We're looking for analysts who have built and validated methods for hazardous compounds — people who know what a detection limit really means and what it implies. What matters most is being able to tell a routine professional question from one that's fishing for something else: the same detection method that finds a compound describes how to evade the screen. You will be writing prompts that sit exactly on that line, then judging whether the model held it. Examples of relevant backgrounds (ideally with red-teaming experience): - Detection and identification of hazardous compounds — GC-MS, LC-MS/MS, IR, NMR - Method development and validation for trace-level analysis - Environmental, occupational or emergency-response sampling and analysis - Reference standards, spectral libraries and unknown identification - Laboratory accreditation, QA/QC and proficiency testing What this role demands beyond technical depth This is writing-intensive work. Every judgment you make needs a written rationale that a non-specialist can follow. Prior technical writing, published research, or expert witness experience is a strong signal; please include a sample or link. You will also be reading and writing about misuse scenarios in your field for sustained periods. We brief experts on this in advance, and you can pause or step away at any point without penalty. Before you apply Your work here will not involve, and must not draw on, classified or export-controlled information, or anything covered by an NDA or prepublication review obligation. If you hold such obligations you may still be a good fit; tell us in your application and we will scope the work accordingly.

Mercor

Audiobook QA Expert — English (US)

Mercor42 days ago

$20-$25/hr

Not specified

Apply

About this role

Help make the next generation of AI-narrated audiobooks sound genuinely human. 1. Overview We're building a team of native English speakers who love audiobooks to help evaluate AI-narrated (text-to-speech) audiobooks in US / North American English. You'll listen the way a real audiobook fan does — but with a careful, analytical ear — and flag the moments where the narration slips, so the underlying models can be made better. This is remote, part-time work at roughly 20 hours per week (about 4 hours per day). You can be based anywhere, though we prefer people living in a region where English is natively spoken. 2. What You'll Do - Listen to AI-narrated audiobooks in English and assess how natural, accurate, and enjoyable the narration is. - Spot and log errors in the synthesized speech — skipped, added, or mispronounced words; numbers read incorrectly; abbreviations expanded wrongly; unnatural phrasing or intonation; audio glitches and artifacts. - Categorize each issue and mark the exact span where it occurs, using our review tooling. - Report on overall listener satisfaction and where the narration feels natural versus where it breaks the experience. - Work consistently and carefully across many hours of audio without losing attention to detail. 3. Who We're Looking For - Native or near-native fluency in English (US / North American English), and an active, genuine audiobook listener — someone who listens regularly and has real opinions about what makes a good listen. - Enough working English to report clearly and communicate with the team. - Some experience with annotation, labeling, transcription, proofreading, linguistics, or other detail-oriented review work. - A sharp, patient ear and the discipline to stay accurate over long listening sessions. - Reliable availability of about 20 hours per week on a part-time basis. Nice to Have - Prior experience on AI training-data, model-evaluation, or audio/QA annotation projects. - A background in language, editing, voice, or audio production.

Mercor

Public Defenders — Paid Research Study

Mercor42 days ago

$100/hr

Worldwide

Apply

About this role

Help map public defenders' work We are recruiting 100 current or former public defenders for a remote research study through Project Mercator. We want to understand the work you personally did, not an ideal job description. What you will do In the first survey, describe your primary responsibilities, tasks, smaller pieces of work, and which tasks need outputs from other tasks. You can also describe different versions of the same task. Written instructions and help from the study coordinator are provided. No outside research or real work product is required. Wave 1 pays $100 per hour of reviewed survey time, up to 90 minutes ($150 maximum). If you need more time, contact the coordinator before continuing beyond the cap. A later survey will ask selected participants to review a combined task tree and add details. Its scope and compensation will be confirmed separately before you begin. Who should apply - Current or former public defenders with hands-on experience representing indigent clients in the United States. - Able to describe the work you personally performed and the jurisdiction and period in which you practiced. - Prior Mercor project experience is welcome but not required. Experience solely as a prosecutor, law student, legal-content annotator, or lawyer without public-defense experience is not sufficient for this study. If your work was court-appointed or contract public defense, explain that arrangement so we can review the fit. How to participate Apply with your resume and a short description of your public-defense experience and availability. We review applications before sending personal survey links. Applying does not automatically enroll you or start the paid survey. Use your own professional experience, not AI-generated answers. Do not include client names, case numbers, identifying facts, confidential information, or privileged material. We do not need any documents from real cases.

Mercor

Source Security Specialist

Mercor42 days ago

$65-$75/hr

Worldwide

Apply

About this role

About the work Mercor is assembling a panel of radiological safety experts to red-team frontier AI models. The goal is to test whether a model can correctly judge the misuse potential of a technical request — answering legitimate questions fully while refusing genuinely dangerous ones. You will: - Write challenging single-turn prompts in your domain, labelled across three levels: benign, dual-use, and adversarial. - Evaluate the responses that come back against a defined policy standard, and judge whether each was handled correctly. - Write the reference answer — what a correct response looks like, and the technical reasoning for why. This domain lives on the dual-use line more than most. The same isotope underpins cancer therapy and industrial radiography; the same shielding calculation protects a worker and conceals a source. A model that refuses a routine dose question is failing as badly as one that answers a question it shouldn't. Drawing that line reliably is why we need practitioners rather than generalists. Background we're looking for We're looking for people who have assessed how Category 1 and 2 sources actually get stolen, shielded or moved — the security and vulnerability side of a licensed programme, not just its paperwork. What matters most is being able to tell a routine professional question from one that's fishing for something else: the same shielding calculation protects a worker and conceals a source. You will be writing prompts that sit exactly on that line, then judging whether the model held it. Examples of relevant backgrounds (ideally with red-teaming experience): - Security or vulnerability assessment of licensed facilities holding Category 1 and 2 sources - Physical protection design and review — detect, delay, assess and respond; insider threat considerations - Source recovery, consolidation or security-upgrade programmes for disused and orphan sources - Regulatory security inspection and compliance — 10 CFR 37 or national equivalents - Transport security for radioactive material — routing, escort, tracking and chain of custody What this role demands beyond technical depth This is writing-intensive work. Every judgment you make needs a written rationale that a non-specialist can follow. Prior technical writing, published research, or expert witness experience is a strong signal; please include a sample or link. You will also be reading and writing about misuse scenarios in your field for sustained periods. We brief experts on this in advance, and you can pause or step away at any point without penalty. Before you apply Your work here will not involve, and must not draw on, classified or export-controlled information, or anything covered by an NDA or prepublication review obligation. If you hold such obligations you may still be a good fit; tell us in your application and we will scope the work accordingly.

Mercor

Propulsion & Effects Engineer

Mercor42 days ago

$65-$75/hr

Worldwide

Apply

About this role

About the work Mercor is assembling a panel of energetic materials and propulsion experts to red-team frontier AI models. The goal is to test whether a model can correctly judge the misuse potential of a technical request — answering legitimate questions fully while refusing genuinely dangerous ones. You will: - Write challenging single-turn prompts in your domain, labelled across three levels: benign, dual-use, and adversarial. - Evaluate the responses that come back against a defined policy standard, and judge whether each was handled correctly. - Write the reference answer — what a correct response looks like, and the technical reasoning for why. This domain lives on the dual-use line more than most. The same propulsion analysis that qualifies a motor for flight describes a capability; the same safe-and-arm logic that keeps a system inert is the part someone would want to defeat. A model that refuses a routine performance or qualification question is failing as badly as one that answers a question it shouldn't. Drawing that line reliably is why we need practitioners rather than generalists. Background we're looking for This is the engineering half of the domain: we're looking for people who have designed, qualified or tested propulsion and initiation systems, with real hardware and real test data behind them. What matters most is being able to tell a routine professional question from one that's fishing for something else: the same effects model that sizes a safety standoff also sizes an intent. You will be writing prompts that sit exactly on that line, then judging whether the model held it. Examples of relevant backgrounds (ideally with red-teaming experience): - Munitions or warhead design and effects engineering - Solid, liquid or hybrid propulsion system design, test and qualification - Terminal effects modelling — blast, fragmentation, penetration - Fuzing, safe-and-arm and initiation system engineering - Test range engineering and instrumentation for propulsion or ordnance What this role demands beyond technical depth This is writing-intensive work. Every judgment you make needs a written rationale that a non-specialist can follow. Prior technical writing, published research, or expert witness experience is a strong signal; please include a sample or link. You will also be reading and writing about misuse scenarios in your field for sustained periods. We brief experts on this in advance, and you can pause or step away at any point without penalty. Before you apply Your work here will not involve, and must not draw on, classified or export-controlled information, or anything covered by an NDA or prepublication review obligation. If you hold such obligations you may still be a good fit; tell us in your application and we will scope the work accordingly.

Mercor

Nuclear Medicine Physicist / Cyclotron RSO

MercorSTEM / Healthcare42 days ago

$65-$75/hr

Worldwide

Apply

About this role

About the work Mercor is assembling a panel of radiological safety experts to red-team frontier AI models. The goal is to test whether a model can correctly judge the misuse potential of a technical request — answering legitimate questions fully while refusing genuinely dangerous ones. You will: - Write challenging single-turn prompts in your domain, labelled across three levels: benign, dual-use, and adversarial. - Evaluate the responses that come back against a defined policy standard, and judge whether each was handled correctly. - Write the reference answer — what a correct response looks like, and the technical reasoning for why. This domain lives on the dual-use line more than most. The same isotope underpins cancer therapy and industrial radiography; the same shielding calculation protects a worker and conceals a source. A model that refuses a routine dose question is failing as badly as one that answers a question it shouldn't. Drawing that line reliably is why we need practitioners rather than generalists. Background we're looking for We're looking for people who produce and handle medical isotopes — the legitimate-use half of the dual-use line, held as a real custody responsibility. What matters most is being able to tell a routine professional question from one that's fishing for something else: the same activity and shielding figures that let a patient go home describe a source someone else wants to move quietly. You will be writing prompts that sit exactly on that line, then judging whether the model held it. Examples of relevant backgrounds (ideally with red-teaming experience): - Nuclear medicine or radiopharmacy physicist — therapy dosimetry, I-131, Lu-177 and Y-90 work, patient release criteria - Cyclotron or radiochemistry facility RSO — target handling, PET isotope production, hot-cell operations - Medical physicist with brachytherapy source custody — HDR and LDR sources, inventory and source exchange - Radiopharmaceutical production and QC under GMP or an equivalent quality system - Authorised Medical Physicist or Authorised User support on a medical radioactive materials licence What this role demands beyond technical depth This is writing-intensive work. Every judgment you make needs a written rationale that a non-specialist can follow. Prior technical writing, published research, or expert witness experience is a strong signal; please include a sample or link. You will also be reading and writing about misuse scenarios in your field for sustained periods. We brief experts on this in advance, and you can pause or step away at any point without penalty. Before you apply Your work here will not involve, and must not draw on, classified or export-controlled information, or anything covered by an NDA or prepublication review obligation. If you hold such obligations you may still be a good fit; tell us in your application and we will scope the work accordingly.

Mercor

Nonproliferation Analyst

Mercor42 days ago

$65-$75/hr

Worldwide

Apply

About this role

About the work Mercor is assembling a panel of nuclear materials and safeguards experts to red-team frontier AI models. The goal is to test whether a model can correctly judge the misuse potential of a technical request — answering legitimate questions fully while refusing genuinely dangerous ones. You will: - Write challenging single-turn prompts in your domain, labelled across three levels: benign, dual-use, and adversarial. - Evaluate the responses that come back against a defined policy standard, and judge whether each was handled correctly. - Write the reference answer — what a correct response looks like, and the technical reasoning for why. This domain lives on the dual-use line more than most. The same control list that stops a shipment tells a procurement agent what to avoid declaring; the same pathway analysis that informs policy is a roadmap read from the other direction. A model that refuses a routine export-classification question is failing as badly as one that answers a question it shouldn't. Drawing that line reliably is why we need practitioners rather than generalists. Background we're looking for We're looking for people who have worked proliferation questions professionally — export control, treaty implementation, or programme analysis — and who can tell a compliance question from a circumvention question. What matters most is being able to tell a routine professional question from one that's fishing for something else: the same control list that blocks a transfer also names what to route around. You will be writing prompts that sit exactly on that line, then judging whether the model held it. Examples of relevant backgrounds (ideally with red-teaming experience): - Export control licensing, classification or enforcement — dual-use lists, catch-all controls, end-user screening - Treaty and agreement implementation — NPT, CTBT, safeguards agreements, bilateral cooperation agreements - Proliferation pathway and programme analysis at a laboratory, agency, ministry or research institute - Sanctions, procurement network or illicit trade analysis - Open-source and technical analysis of nuclear programmes What this role demands beyond technical depth This is writing-intensive work. Every judgment you make needs a written rationale that a non-specialist can follow. Prior technical writing, published research, or expert witness experience is a strong signal; please include a sample or link. You will also be reading and writing about misuse scenarios in your field for sustained periods. We brief experts on this in advance, and you can pause or step away at any point without penalty. Before you apply Your work here will not involve, and must not draw on, classified or export-controlled information, or anything covered by an NDA or prepublication review obligation. If you hold such obligations you may still be a good fit; tell us in your application and we will scope the work accordingly.

Mercor

Specialty Materials Chemist — Pyrotechnics or Propulsion

MercorSTEM42 days ago

$65-$75/hr

Worldwide

Apply

About this role

About the work Mercor is assembling a panel of energetic materials and propulsion experts to red-team frontier AI models. The goal is to test whether a model can correctly judge the misuse potential of a technical request — answering legitimate questions fully while refusing genuinely dangerous ones. You will: - Write challenging single-turn prompts in your domain, labelled across three levels: benign, dual-use, and adversarial. - Evaluate the responses that come back against a defined policy standard, and judge whether each was handled correctly. - Write the reference answer — what a correct response looks like, and the technical reasoning for why. This domain lives on the dual-use line more than most. The same formulation work that qualifies a propellant or a display composition sits next to something nobody should be helped with; the same sensitivity data that protects a production line reads as a specification from the other side. A model that refuses a routine thermal-stability question is failing as badly as one that answers a question it shouldn't. Drawing that line reliably is why we need practitioners rather than generalists. Background we're looking for This is the chemistry half of the domain: we're looking for people who have formulated, synthesised and tested specialty energetic materials in a licensed or government facility, not modelled them from a distance. What matters most is being able to tell a routine professional question from one that's fishing for something else: the same sensitivity result that sets a safe handling limit is what someone else would want to design around. You will be writing prompts that sit exactly on that line, then judging whether the model held it. Examples of relevant backgrounds (ideally with red-teaming experience): - Formulation chemist for pyrotechnics, propellants or specialty energetic materials - Synthesis and characterisation of energetic compounds in a licensed or government facility - Performance and sensitivity testing — calorimetry, thermal stability, small-scale safety testing - Scale-up and process safety for energetic material production - Solid or liquid rocket propellant development — binders, oxidisers, burn-rate modifiers What this role demands beyond technical depth This is writing-intensive work. Every judgment you make needs a written rationale that a non-specialist can follow. Prior technical writing, published research, or expert witness experience is a strong signal; please include a sample or link. You will also be reading and writing about misuse scenarios in your field for sustained periods. We brief experts on this in advance, and you can pause or step away at any point without penalty. Before you apply Your work here will not involve, and must not draw on, classified or export-controlled information, or anything covered by an NDA or prepublication review obligation. If you hold such obligations you may still be a good fit; tell us in your application and we will scope the work accordingly.

Mercor

Audiobook QA Expert — Italian

MercorLanguages42 days ago

$15-$20/hr

Not specified

Apply

About this role

Help make the next generation of AI-narrated audiobooks sound genuinely human. 1. Overview We're building a team of native Italian speakers who love audiobooks to help evaluate AI-narrated (text-to-speech) audiobooks in Italian (Italy). You'll listen the way a real audiobook fan does — but with a careful, analytical ear — and flag the moments where the narration slips, so the underlying models can be made better. This is remote, part-time work at roughly 20 hours per week (about 4 hours per day). You can be based anywhere, though we prefer people living in a region where Italian is natively spoken. 2. What You'll Do - Listen to AI-narrated audiobooks in Italian and assess how natural, accurate, and enjoyable the narration is. - Spot and log errors in the synthesized speech — skipped, added, or mispronounced words; numbers read incorrectly; abbreviations expanded wrongly; unnatural phrasing or intonation; audio glitches and artifacts. - Categorize each issue and mark the exact span where it occurs, using our review tooling. - Report on overall listener satisfaction and where the narration feels natural versus where it breaks the experience. - Work consistently and carefully across many hours of audio without losing attention to detail. 3. Who We're Looking For - Native or near-native fluency in Italian (Italy), and an active, genuine audiobook listener — someone who listens regularly and has real opinions about what makes a good listen. - Enough working English to report clearly and communicate with the team (you do not need to be bilingual). - Some experience with annotation, labeling, transcription, proofreading, linguistics, or other detail-oriented review work. - A sharp, patient ear and the discipline to stay accurate over long listening sessions. - Reliable availability of about 20 hours per week on a part-time basis. Nice to Have - Prior experience on AI training-data, model-evaluation, or audio/QA annotation projects. - A background in language, editing, voice, or audio production.

Mercor

Industrial Hygienist / Process Safety Engineer

Mercor42 days ago

$65-$75/hr

Worldwide

Apply

About this role

About the work Mercor is assembling a panel of chemistry and chemical safety experts to red-team frontier AI models. The goal is to test whether a model can correctly judge the misuse potential of a technical request — answering legitimate questions fully while refusing genuinely dangerous ones. You will: - Write challenging single-turn prompts in your domain, labelled across three levels: benign, dual-use, and adversarial. - Evaluate the responses that come back against a defined policy standard, and judge whether each was handled correctly. - Write the reference answer — what a correct response looks like, and the technical reasoning for why. This domain lives on the dual-use line more than most. The same dispersion model that sites a plant and sets an evacuation radius also predicts an exposure; the same hazard analysis that finds a failure path is a list of failure paths. A model that refuses a routine exposure-limit question is failing as badly as one that answers a question it shouldn't. Drawing that line reliably is why we need practitioners rather than generalists. Background we're looking for We're looking for people who have assessed real exposures and real process hazards, and signed off on the controls. What matters most is being able to tell a routine professional question from one that's fishing for something else: the same release model that protects a community estimates a consequence. You will be writing prompts that sit exactly on that line, then judging whether the model held it. Examples of relevant backgrounds (ideally with red-teaming experience): - Certified industrial hygienist — exposure assessment, sampling strategy, control banding - Process hazard analysis — HAZOP, LOPA, what-if studies, management of change - Toxic release and dispersion modelling — consequence analysis, AEGL/ERPG criteria, facility siting - Ventilation, containment and PPE programme design for chemical hazards - Regulatory compliance — PSM, Seveso, COMAH, OSHA or national equivalents What this role demands beyond technical depth This is writing-intensive work. Every judgment you make needs a written rationale that a non-specialist can follow. Prior technical writing, published research, or expert witness experience is a strong signal; please include a sample or link. You will also be reading and writing about misuse scenarios in your field for sustained periods. We brief experts on this in advance, and you can pause or step away at any point without penalty. Before you apply Your work here will not involve, and must not draw on, classified or export-controlled information, or anything covered by an NDA or prepublication review obligation. If you hold such obligations you may still be a good fit; tell us in your application and we will scope the work accordingly.

Mercor

Federal Certified Materials Specialist / Incident Analyst

Mercor42 days ago

$65-$75/hr

Worldwide

Apply

About this role

About the work Mercor is assembling a panel of energetic materials and propulsion experts to red-team frontier AI models. The goal is to test whether a model can correctly judge the misuse potential of a technical request — answering legitimate questions fully while refusing genuinely dangerous ones. You will: - Write challenging single-turn prompts in your domain, labelled across three levels: benign, dual-use, and adversarial. - Evaluate the responses that come back against a defined policy standard, and judge whether each was handled correctly. - Write the reference answer — what a correct response looks like, and the technical reasoning for why. This domain lives on the dual-use line more than most. The same residue analysis that identifies what was used is a description of what to use; the same diversion indicators that catch a bad licensee tell someone how not to be caught. A model that refuses a routine forensic identification question is failing as badly as one that answers a question it shouldn't. Drawing that line reliably is why we need practitioners rather than generalists. Background we're looking for We're looking for people who have worked scenes and casework under a national certification — identifying materials after the fact, and inspecting the licensees who hold them. What matters most is being able to tell a routine professional question from one that's fishing for something else: the same residue signature that closes a case describes a recipe in reverse. You will be writing prompts that sit exactly on that line, then judging whether the model held it. Examples of relevant backgrounds (ideally with red-teaming experience): - Federally certified explosives specialist or national equivalent — scene and materials analysis - Forensic laboratory identification of energetic materials and post-blast residues - Post-incident analysis and reporting for an agency or laboratory - Regulatory inspection of licensees and permittees — storage, records and diversion indicators - Expert testimony or casework on energetic materials What this role demands beyond technical depth This is writing-intensive work. Every judgment you make needs a written rationale that a non-specialist can follow. Prior technical writing, published research, or expert witness experience is a strong signal; please include a sample or link. You will also be reading and writing about misuse scenarios in your field for sustained periods. We brief experts on this in advance, and you can pause or step away at any point without penalty. Before you apply Your work here will not involve, and must not draw on, classified or export-controlled information, or anything covered by an NDA or prepublication review obligation. If you hold such obligations you may still be a good fit; tell us in your application and we will scope the work accordingly.

Mercor AI training jobs. Earn on your schedule.

Earnings Calculator

Filters

Job listings281

About this role

About this role

About this role

About this role

About this role

About this role

About this role

About this role

About this role

About this role

About this role

About this role

About this role

About this role

About this role

About this role

About this role

About this role

About this role

About this role