Ethos needs its first dedicated compliance hire to build the foundations for financial crime controls and data privacy. You’ll manage KYC, AML, sanctions, fraud and privacy work hands-on, investigate flagged activity, respond to data and compliance requests, and write the policies that guide decisions. The role is based at the London HQ, with office work Monday to Thursday and remote work on Friday.
Help assess how AI models respond to safety-critical questions about radioactive materials and health physics. You will create evaluation prompts, judge work from other experts, and explain the technical basis for each safety decision. The work is remote and task-based, and you must be a citizen of and work from a Five Eyes country.
Record short German audio clips and perform brief scripts with varied tones, emotions, and speaking styles. You may also review recordings, capture details in text, and help AI systems interpret speech and vocal expression. Native German fluency and comfort speaking into a microphone matter more than formal acting credentials.
Review AI-generated physics answers for sound reasoning, correct calculations, and clear explanations. You’ll write prompts and model solutions, check claims against public sources, and compare responses across topics such as mechanics, electromagnetism, and optics. Applicants need a relevant graduate degree from a top 100 university, peer-reviewed publication experience, and C1-level English.
A paid study will use practicing economists’ analysis to strengthen AI performance on complex economic questions. You’ll help build benchmarks in forecasting, antitrust, and trade policy, contributing to more accurate analytical tools. The work is remote and asynchronous, with a commitment of at least 20 hours per week. Applicants must be based in an eligible country.
Handshake AI’s year-round fellowship brings STEM researchers into project-based work that helps improve large language models across scientific and technical fields. You’ll create domain-specific prompts, assess model answers, and research topics with AI support. Work takes place remotely on an asynchronous schedule, with a commitment of about 10 to 20 hours per week. Project placement depends on availability in your discipline.
Use your production SOC experience to review alerts, verify investigation evidence, and assess whether security conclusions hold up. The work spans SIEM, endpoint, cloud, and identity environments, with hands-on Splunk analysis at its core. Some assignments call for a clear ACCEPT or PASS decision; others require a complete investigation with documented evidence and reasoning.
Heidi Health is collecting short speech recordings from UK-based clinicians to train and test its speech recognition system across regional and community accents. You’ll record one clinical passage and one non-medical passage, a total of about 15 minutes, using a phone or laptop microphone. Heidi checks an unpaid sample for accent match and audio quality before any paid recording; the sample does not train a model.
Apply your expertise in explosives or energetic materials to assess how AI models handle safety-sensitive questions. You will create and review prompts that test the line between acceptable and unsafe requests, then document the technical basis for each decision. Applicants must hold a relevant master's degree or higher, be citizens of a Five Eyes country, and live in one of those countries while doing the work.
Record short clips and read brief scripts with varied tones, emotions, and speaking styles. You may also review audio, document what you hear, and help AI systems interpret speech and expression. The project pays up to $10 USD per hour for core work; Missions may offer extra incentives, and rates can vary. Some onboarding, assessment, or non-core work may have a different rate.
Build realistic Rhino modeling challenges that reflect professional design work, then document their context in SuperAnnotate and record your workflow. Tasks can cover NURBS surfaces, solids, geometry repair, architecture, or product design. You will prepare the software and project assets, and check model quality and export integrity before submitting your recordings and files.
Take part in a paid, AI-moderated interview about your experience with everyday apps and websites. You’ll answer questions about your digital habits, try a few simple tasks on your own device, and share what makes interfaces easy or hard to use. The study welcomes consumers from any professional background who use digital tools on a regular basis.
Use your experience in computational, physical, or theoretical chemistry to create expert-level problems and assess AI-generated answers. You’ll check technical accuracy, reproducibility, and whether responses reflect real chemistry workflows, then provide written feedback. The work is ongoing and project-based, with a flexible schedule and no minimum hour requirement. Applicants need UK work authorization.
Pearson is building an ongoing pool of practising lawyers to review AI responses to real legal questions across several jurisdictions. You will assess legal accuracy, reasoning, and completeness, then write feedback under project-specific guidelines. Work takes place remotely, starts at around 10 hours with potential for more, and falls under an NDA. The team reviews applications on a rolling basis as projects arise.
At Ethos, you’ll handle support for experts whose work helps AI labs and other frontier clients build better models. You’ll own each case from the first message through resolution, work with internal teams, and review AI-drafted replies for accuracy and empathy. The role is based at the London office near King’s Cross, with in-person work Monday through Thursday and remote work on Friday.
Help assess how AI models handle safety-sensitive questions in materials science. You will create evaluation prompts, review work from other experts, and explain the technical basis for each safe or unsafe classification. The role requires a graduate degree in a relevant discipline and citizenship and residence in a Five Eyes country while you complete the work.
Record short Korean audio and complete related text tasks that help AI systems recognize human speech, tone, and expression. Work may include reading or performing scripts with varied emotions and styles, and reviewing audio or video clips to document details. Core project work pays up to $31 USD per hour; rates can vary, and some onboarding, assessment, or non-core work may pay at a different rate.
Review and create Mandarin content for AI training, judging response quality, reasoning, and fidelity to the prompt. Apply China-specific localization expertise and MQM/LQA methods to identify language and factual issues, compare model answers, and write sound reference explanations. The work calls for native or near-native Mandarin, C1 or higher English, and at least five years of localization or translation experience.
Contribute to a paid, ongoing project that assesses how accounting professionals use QuickBooks for bookkeeping and financial reporting. You will work remotely for 30 to 40 hours each week, handle accounting tasks in the software, and report on workflow efficiency, bottlenecks, and possible improvements. The project welcomes professionals with experience in small business or enterprise accounting.
Record voice samples for a text-to-speech system that supports a client’s internal CX AI Agent. The client seeks women based in northern UK who speak with an authentic Northern English accent, including Yorkshire, Manchester, Newcastle, or Liverpool varieties. The work involves one recording session of about four hours, with a possible second session of similar length. The client will not sell, license, or reuse recordings for another purpose.
Location and eligibility
Remote · United Kingdom only
Engagement type
Hourly
Compensation
$50–$100/hr
voice acting
voice cloning
text-to-speech
AI trainingView job
Showing 20 active roles of 81. Dates reflect what the source published, not when AIWorkJobs last synced.
Sources: ten specialist AI work platforms. AIWorkJobs does not verify every claim made by a source. Report a listing that looks wrong through the contact page.