Regularly reviewed
Listings are checked on a regular schedule. A review date is not a guarantee that a role remains open between checks.
Verified remote AI jobs, data annotation gigs, RLHF work & freelance AI opportunities from Scale AI, Outlier, Mercor, Surge AI, Turing & more — independently curated for anyone ready to work with AI.
Pay figures shown anywhere on this site are rates advertised by the companies themselves. They are not guaranteed, represent the top of a range rather than a typical rate, and vary by role, country, and assessment results. HowToAIjob does not set or pay these rates.
CURATED DIRECTORY — review eligibility and current terms on the official page · compare every advertised rate in the Rate Index · new here? how hiring works ↓
FullStack Connect presents remote software and tech opportunities across multiple countries. The official talent page displays examples such as React Native Developer, Python Developer, QA Automation Engineer and Product Designer, with both mid-level and senior experience levels. Its five-step process covers a phone screen, asynchronous skill assessment and interview, profile creation and client matching. Checked 28 Oct 2026.
Read before applyingOfficial talent processReading contractor payCompare all 49 platforms
Official talent and Connect pages, 28 Oct 2026 · worldwide label updated after job-board reviewBrowse FullStack Connect jobs ↗Ethos matches experts with consulting, advisory, research and other paid opportunities. Its official opportunity page currently lists a fully remote Market Research & Consumer Insights Manager project supporting a leading AI lab's presentation and slide-deck work. The role advertises $70/hr for 5–20 hours per week and asks for 4+ years of relevant experience; an AI voice screen may be part of the application.
SourcesOfficial $70/hr opportunitya16z investment announcementReading contractor pay
Official Ethos opportunity page + a16z investment announcement · checked 24 Oct 2026Join Ethos via official invite ↗Workada's remote Data Labeling Specialist contract involves evaluating AI outputs across text, images, video and documents against detailed criteria, spotting errors and inconsistencies, and writing clear justifications. The official opportunities page advertises $17/hr for this role. Workada says training is paid and, after setup, Stripe payouts are sent three times a week; the first payout can take extra business days.
Read before applyingWhat annotation work actually involvesReading an advertised rate as a contractorCheck this rate against all 49 platforms
Verified on Workada official role and opportunities pages, 26 Oct 2026Apply on workada.com ↗Terac's official homepage lists 420,000+ registered experts, 500+ open opportunities and 130+ countries. Its six reviewed expert tracks advertise remote AI-review work at $50–$200/hr: accounting $50–$140, healthcare operations $50–$120, finance $60–$160, software engineering $60–$200, management consulting $90–$200 and product/platform $70–$160. Terac says applicants can apply once for rolling matches and do not need prior AI experience; matching and pay depend on expertise and project. Checked 3 Nov 2026.
Read before applyingReading an advertised rate as a contractorCheck this rate against all 49 platformsExpert & RLHF role tracks
Official homepage + six expert tracks · checked 3 Nov 2026Apply on terac.com ↗CloudFactory's official Workable board displayed 29 jobs during the 3 Nov 2026 review. Visible examples include Bilingual Computational Linguist, Forward Deployed Engineer, AI Marketing Operations Manager, Client Principal - AI Consulting, and construction data/reality-capture technician roles. The board includes remote, hybrid and on-site work, with full-time, contract and part-time openings; location filters can change the visible results.
Read before applyingExpert & RLHF role tracksHow to apply to remote AI rolesRed flags worth 60 seconds before you send anything
Official Workable board · checked 3 Nov 2026Apply on cloudfactory.com ↗CVAT.ai's official careers page lists six openings: Backend Developer, Technical Writer, DevOps Engineer, Senior Python Developer, Software Engineer in Test and Data Annotator. The first five are full-time and listed as Remote Worldwide or Cyprus; Data Annotator is a contract role listed as Remote (Worldwide). Checked 3 Nov 2026.
Read before applyingEvery advertised rate in one tableHow to apply to remote AI rolesRed flags worth 60 seconds before you send anything
Official careers + Data Annotator pages · checked 3 Nov 2026Apply on cvat.ai ↗Anthropic's official careers and jobs pages list opportunities across AI research and engineering, policy, operations and business. Role locations include San Francisco, New York, Seattle, London, Zürich and Ontario, with office-based, hybrid and selected remote-friendly arrangements. The careers page advertises competitive salary/equity and a $500/month flexible wellness and time-saver stipend; eligible roles can offer visa and green-card sponsorship. Checked 3 Nov 2026.
Read before applyingEvery advertised rate in one tableExpert & RLHF role tracksHow to apply to remote AI roles
Official careers + jobs pages · checked 3 Nov 2026Apply on anthropic.com ↗Shaip's official community page lists 500,000+ contributors, $7.5M+ total earnings, 70+ countries and project categories totalling 2,400+ advertised projects. It advertises weekly payouts and says many tasks need no technical experience. The current Indic audio project accepts eligible existing audio content in Hindi and other Indic languages; its project-specific rate is up to ₹200 per audio hour, not an hourly-work guarantee. Checked 29 Oct 2026.
Read before applyingTaxes, payouts and withdrawal feesRed flags worth 60 seconds before you send anythingEvery advertised rate in one table
Official community + Indic project pages · checked 29 Oct 2026Apply on shaip.comSurge's official workforce page currently lists seven 100% remote contractor roles for constitutional lawyers, journalists, management consultants, investment bankers, VC partners, startup CEOs and professors. Current official role results for selected lawyer, journalist, consultant and banker openings advertise $200–$400/hour. Apply by emailing the recruiting contact on the official page with your background and role name. Checked 29 Oct 2026.
Read before applyingReading an advertised rate as a contractorCheck this rate against all 49 platformsExpert & RLHF role tracks
Official workforce + role results · checked 29 Oct 2026Apply on surgehq.aiTether's official careers page listed 208 jobs on 29 Oct 2026 across Compliance, Engineering, Expansion, Finance, Legal, Marketing, Product, and Regulatory Affairs & Licensing. Current role pages include remote AI research and engineering openings and India locations including Bangalore; eligibility is role-specific. Apply only through the official careers page.
Read before applyingHow to apply to remote AI rolesRed flags worth 60 seconds before you send anythingEvery advertised rate in one table
Official careers + role pages · checked 29 Oct 2026Apply on tether.ioElevenLabs' official careers pages listed 153 open positions across 52 locations on 29 Oct 2026. Current openings include full-time engineering, growth, revenue and operations plus freelance audiobook, dubbing, transcription/subtitling, translator/linguist and audio-engineering roles. The company says its team spans 50+ countries and supports remote work, but its setup is hybrid and role location varies. An India-specific Deployment Strategist Lead opening was also listed.
Read before applyingCheck this rate against all 49 platformsWhat annotation work actually involvesExpert & RLHF role tracks
Official careers + positions pages · checked 29 Oct 2026Apply on elevenlabs.ioScholars' current site lists 18 roles across healthcare, STEM, languages, humanities and other fields. It says tasks pay hourly, sign-up is free and open to people of legal age, and skill assessments are used for matching. Students and researchers can apply, but becoming a Scholar does not guarantee contact, acceptance or work. Checked 29 Oct 2026.
Read before applyingWhat annotation work actually involvesExpert & RLHF role tracksRed flags worth 60 seconds before you send anything
Official Scholars site · checked 29 Oct 2026Apply on scholars.netAI Acquisition's official Workable board displayed 17 jobs during the 3 Nov 2026 review, including Client Onboarding Concierge, B2B Sales Representative, Account Manager, Senior Strategic Growth Partner, Community Growth Manager and Direct Response Copywriter. Visible remote roles span multiple countries, including the UK, US, Canada, South Africa, Poland, Brazil and Egypt; eligible locations differ by job. This is an internal corporate job board, not an AI task marketplace.
Read before applyingHow to apply to remote AI rolesRed flags worth 60 seconds before you send anythingEvery advertised rate in one table
Official Workable board · checked 3 Nov 2026Apply on aiacquisition.com ↗Snorkel AI offers two genuinely different paths: full-time corporate roles in Engineering, Research, Sales and Product (mostly based in San Francisco/Redwood City, hybrid), and a separate Expert Community for PhDs and domain specialists who want contract-based, remote-friendly work building AI training datasets. Pick whichever matches what you're looking for.
Read before applyingReading an advertised rate as a contractorCheck this rate against all 49 platformsExpert & RLHF role tracks
Apply on snorkel.ai ↗Avalara is a publicly-traded (NYSE: AVLR) tax compliance and AI platform with 200,000+ customers across 75+ countries. This is a stable, traditional corporate job rather than a freelance gig — Glassdoor shows a 3.2/5 employee rating with 49% recommending it to a friend, which is solid but not exceptional. Open roles include account management, benefits systems, and engineering positions, with some based in Pune, India.
Read before applyingEvery advertised rate in one tableHow to apply to remote AI rolesRed flags worth 60 seconds before you send anything
Apply on avalara.com ↗Besimple AI is a Y Combinator-backed startup building the data layer for voice AI. Their contributor side, Pila8, pays everyday people (not just engineers) to record short voice clips that train AI to understand different accents and languages — base rate is $0.10/minute, with non-English speakers and consistent contributors earning bonuses.
Read before applyingWhat annotation work actually involvesRed flags worth 60 seconds before you send anythingEvery advertised rate in one table
Apply on besimple.com ↗DataForce is TransPerfect's AI data platform with over 1.3 million remote contributors worldwide, working on data annotation, transcription, search evaluation, and AI training tasks. It's a legitimate, established side of TransPerfect (25+ years, 140+ cities) — but reviews say to treat it as flexible freelance income, not a stable full-time job, since project availability varies by location and language demand.
Read before applyingCheck this rate against all 49 platformsWhat annotation work actually involvesStarting from zero: the beginner guide
Apply on dataforce.ai ↗Revolut is a global fintech "super app" with 60+ million customers and 622 open roles across Engineering, Compliance, Sales, Finance, and Legal. It's genuinely remote-first — 239 of its open roles are fully remote — and known for fast growth, strong pay (engineers average $126K+), and an intense, results-driven culture rather than a relaxed one.
Read before applyingEvery advertised rate in one tableHow to apply to remote AI rolesRed flags worth 60 seconds before you send anything
Apply on revolut.com ↗Welo Global is a 25+ year old language and localization company trusted by Google, Spotify, Amazon, and other major brands — they handle translation, transcription, data annotation, and AI training tasks across 300+ languages. It's one of the more established names in this space, which generally means more structured onboarding and steadier project flow compared to smaller gig platforms.
Read before applyingStarting from zero: the beginner guideRed flags worth 60 seconds before you send anythingEvery advertised rate in one table
Apply on welo.com ↗Crossover has run remote, performance-based hiring for years, placing people in roles with Fortune 500 clients across 100 countries, with a wide pay band reflecting very different role levels. The "top 1%" framing means the screening process is intentionally selective.
Read before applyingRed flags worth 60 seconds before you send anythingEvery advertised rate in one tableHow to apply to remote AI roles
Apply on crossover.com ↗GitLab is a major, publicly traded (Nasdaq: GTLB) DevOps platform company that's run fully remote since it was founded — no offices, no headquarters, just a globally distributed team across 65+ countries. It's a different kind of listing than most others here: this is full-time employment at an established public tech company, not freelance or gig-based AI training work.
Read before applyingEvery advertised rate in one tableHow to apply to remote AI rolesRed flags worth 60 seconds before you send anything
Apply on gitlab.com ↗Built on Handshake's existing university network, this fellowship targets PhD and Master's-level fellows who evaluate or generate advanced AI responses in their academic specialty. It's one of the higher-paying options here, but it leans heavily on formal qualifications.
Read before applyingRed flags worth 60 seconds before you send anythingEvery advertised rate in one tableReading an advertised rate as a contractor
Apply on joinhandshake.com ↗Pareto AI runs a human-data platform that connects AI labs with vetted subject-matter experts for RLHF, prompt evaluation, and domain-specific annotation — across fields like medicine, law, finance, and engineering. It sits at the high end of this list: the company says general work pays $35–$60/hr, with specialized expert tasks going up to $230/hr, and that it has raised $37M in funding so far.
Read before applyingEvery advertised rate in one tableReading an advertised rate as a contractorCheck this rate against all 49 platforms
Apply on pareto.ai ↗Rex.zone connects subject-matter experts with AI labs that need domain-specific data — model answers get checked, scored, or rewritten across fields like law, medicine, finance and more. Work is task-based and remote, so you pick projects that match what you actually know.
Read before applyingReading an advertised rate as a contractorCheck this rate against all 49 platformsExpert & RLHF role tracks
Apply on rex.zone ↗Scale AI is one of the most established names in AI training data, founded in 2016 and used by many major AI labs. As the most well-known platform on this list, expect higher applicant volume and correspondingly more competitive task qualification.
Read before applyingCheck this rate against all 49 platformsRed flags worth 60 seconds before you send anythingEvery advertised rate in one table
Apply on scale.com ↗Stellar AI is a freelance AI training platform open to contributors worldwide with no prior experience required — you complete tasks like rating responses, writing prompts, or evaluating AI outputs, and get paid hourly. The "$25/hr+" framing means that's the starting rate, with higher-paying tasks available depending on skill and task type. Being "open-ended contract" means there's no fixed end date, but also no guaranteed hours — work is available when projects are live.
Read before applyingStarting from zero: the beginner guideRed flags worth 60 seconds before you send anythingEvery advertised rate in one table
Apply on stellarai.com ↗Alignerr recruits ML experts to label and validate robotics-related training data. Worth flagging clearly: the $1000 shown on this card is a referral bonus for bringing in other qualified experts, not the hourly pay for the work itself — actual task pay is set separately once you're onboarded.
Read before applyingRed flags worth 60 seconds before you send anythingEvery advertised rate in one tableReading an advertised rate as a contractor
Apply on alignerr.com ↗AfterQuery is a YC-backed startup hiring AI experts to evaluate and refine model outputs across 40+ open roles. Being YC-backed means it went through outside investor vetting, which is a decent — though not guaranteed — signal of legitimacy.
Read before applyingEvery advertised rate in one tableReading an advertised rate as a contractorCheck this rate against all 49 platforms
Apply on afterquery.com ↗DataAnnotation advertised 112 roles on 28 Oct 2026 across General, Coding, Languages, STEM and professional tracks. Its current FAQ lists General projects at $25–$50/hr and Coding at $40–$150+/hr; rates and requirements vary by qualification track.
Read before applyingCheck this rate against all 49 platformsRed flags worth 60 seconds before you send anythingEvery advertised rate in one table
Apply on dataannotation.tech ↗An invite-only network for top-tier AI experts. Invite-only models usually mean a higher entry bar and limited slots, so this one's less about applying on demand and more about getting noticed by the network in the first place.
Read before applyingExpert & RLHF role tracksStarting from zero: the beginner guideRed flags worth 60 seconds before you send anything
Apply on deccanai.com ↗Innodata is a publicly traded (Nasdaq) AI data engineering company that's been operating for decades, working with major tech clients. As a listed public company, it has more public financial transparency than most platforms on this page.
Read before applyingEvery advertised rate in one tableWhat annotation work actually involvesHow to apply to remote AI roles
Apply on innodata.com ↗Focused specifically on language and audio work — transcription, voice recording, pronunciation review for AI speech models. It's a narrower niche than most listings here, so it suits people with strong audio or language skills more than general data work.
Read before applyingRed flags worth 60 seconds before you send anythingEvery advertised rate in one tableReading an advertised rate as a contractor
Apply on linguasenseai.com ↗Similar in model to Mercor and Turing — micro1 matches AI experts with frontier lab projects, with pay scaling by skill level. A few platforms on this page use this same "staffing agency for AI labs" approach, so requirements tend to overlap.
Read before applyingReading an advertised rate as a contractorCheck this rate against all 49 platformsExpert & RLHF role tracks
Apply on micro1.ai ↗Mindrift runs AI training projects across many skill areas and languages, with pay scaling from entry-level to expert tasks. It's a broad, multi-skill platform rather than one narrow niche, so most backgrounds can usually find something to apply for.
Read before applyingExpert & RLHF role tracksStarting from zero: the beginner guideRed flags worth 60 seconds before you send anything
Apply on mindrift.ai ↗OneForma focuses on linguistic and data-collection microtasks — translation checks, transcription, search relevance — across 60+ languages. Pay per task is usually modest since tasks are short, but it's accessible if your language isn't widely supported elsewhere.
Read before applyingRed flags worth 60 seconds before you send anythingEvery advertised rate in one tableReading an advertised rate as a contractor
Apply on oneforma.com ↗Outlier runs expert evaluation tasks — reviewing, scoring, and correcting AI-generated answers — across 60+ open roles in subjects like coding, math, and writing. Pay is task-based rather than a flat hourly rate, so it depends on which project you qualify for.
Read before applyingEvery advertised rate in one tableReading an advertised rate as a contractorCheck this rate against all 49 platforms
Apply on outlier.ai ↗Prolific is a bit different from the rest of this list — instead of AI-labeling tasks, it connects you with paid academic and commercial research studies and surveys, some of which feed into AI/UX research. Pay is per-study rather than ongoing hourly task work.
Read before applyingEvery advertised rate in one tableReading an advertised rate as a contractorCheck this rate against all 49 platforms
Apply on prolific.com ↗RWS is an established language-services company expanding into AI training data work through TrainAI, covering translation-adjacent and linguistic annotation tasks. Pay here sits on the lower end of this list, which is typical for high-volume linguistic microtasks.
Read before applyingCheck this rate against all 49 platformsWhat annotation work actually involvesRed flags worth 60 seconds before you send anything
Apply on rws.com ↗A part-time Prompt Engineer role — writing and refining prompts used to test or steer AI model behavior. It's narrower and more specialized than most listings here, and the pay reflects a part-time rather than full freelance commitment.
Read before applyingCheck this rate against all 49 platformsRed flags worth 60 seconds before you send anythingEvery advertised rate in one table
Apply on sarootlabs.com ↗TELUS Digital's AI data arm runs annotation and labeling projects worldwide, currently with seven open contributor roles. Being part of a large, established company, the process tends to be more structured than smaller gig platforms.
Read before applyingRed flags worth 60 seconds before you send anythingEvery advertised rate in one tableReading an advertised rate as a contractor
Apply on telusinternational.com ↗Turing places software and AI engineers with major companies on a project basis, including LLM-training work like agent function-calling — teaching a model to correctly trigger tools and APIs. Engagements are typically several weeks long rather than one-off tasks.
Read before applyingEvery advertised rate in one tableReading an advertised rate as a contractorCheck this rate against all 49 platforms
Apply on turing.com ↗Listed as a flexible freelance contributor role. We don't have much first-hand detail on this one yet, so treat it like any unfamiliar gig listing: read the actual offer page closely before committing real time to it.
Read before applyingReading an advertised rate as a contractorCheck this rate against all 49 platformsRed flags worth 60 seconds before you send anything
Apply on uber.com ↗xAI is running a large recruiting push — 219 listed roles — for people who help train and evaluate Grok. A drive at this scale usually means varied roles at very different pay levels, so "up to $200/hr" applies to specific senior tasks, not every opening.
Read before applyingCheck this rate against all 49 platformsRed flags worth 60 seconds before you send anythingEvery advertised rate in one table
Apply on x.ai ↗Appen is a long-running, publicly listed data company, and CrowdGen is its crowdsourced worker portal for things like search evaluation, data collection, and transcription in 200+ countries. It's a high-volume microtask model — individual tasks pay modestly but availability is broad.
Read before applyingStarting from zero: the beginner guideRed flags worth 60 seconds before you send anythingEvery advertised rate in one table
Apply on appen.com ↗SuperAnnotate's subject-matter-expert track pays a wide range depending on how specialized your knowledge is, with roles open across 40+ countries. The wide $20–$130+/hr band reflects how much subject expertise can change your rate.
Read before applyingRed flags worth 60 seconds before you send anythingEvery advertised rate in one tableReading an advertised rate as a contractor
Apply on superannotate.com ↗Invisible Technologies runs the Meridial Expert Network, an operations-and-data partner for several major AI labs, with a large existing pool of experts already onboarded. Being part of a bigger network can mean steadier project flow, but also more competition to get matched.
Read before applyingEvery advertised rate in one tableReading an advertised rate as a contractorCheck this rate against all 49 platforms
Apply on invisible.co ↗MaxAccelerate is a 25-year-old IBM Maximo and Enterprise Asset Management specialist with 150+ global clients across 170+ countries. They're hiring remote IBM Maximo developers, EAM consultants, sales leads, and part-time recruiters — with an unusual Dubai relocation option offering tax-free earnings and company sponsorship for select roles.
Read before applyingHow to apply to remote AI rolesRed flags worth 60 seconds before you send anythingEvery advertised rate in one table
Apply on maxaccelerate.com ↗Almost none of the 49 listings above is an open sign-up. Each one gates work behind an application and an entry assessment, then releases tasks in project waves. The four stages below are the same on Outlier, DataAnnotation.tech, Mercor, Alignerr, TELUS Digital, Appen and the rest — only the wording and the wait change. The card for each company tells you what is specific to it; this section explains what applies to all of them.
The first step is an account with an accurate background, languages and domain, followed by a qualification test. On most platforms the assessment is unpaid and takes 30–90 minutes. It is scored on consistency and instruction-following far more than on speed: the same rule applied the same way to many similar examples. Read the guideline document end to end before you start answering — most failures are skimming, not ability.
Passing does not create a job. It puts you in a pool, and work is released in waves when a client project opens in your language or domain. An empty queue in week one is normal and is not a rejection. Watch the platform dashboard rather than your inbox — many platforms never send a rejection email, and some never send an acceptance email either.
Before the first payout you will usually be asked for a government ID, address and tax form. Your platform name, ID name and payout account name must match exactly; mismatches are the single most common reason a payment is held. Complete these checks the same week you are approved rather than when the money is due.
Budget two to four weeks from approval to the first payment, then a weekly, bi-weekly or monthly cycle depending on the platform. Payment rails differ — PayPal, Payoneer, Wise, direct transfer or Deel — and the fee is yours to carry. The “Listed” date on each card is the date we last opened that company's page, not a promise that the rate is unchanged today.
The big number on a card is the top of the range the company publishes, not what a new contributor earns in week one. Entry tasks are normally priced one or two tiers below the headline, assessments and waiting time are unpaid, and several crowd platforms adjust pay by country. Read the rate as a ceiling to work towards, then compare it against the payment fees and the unpaid weeks using our contractor pay guide before you decide it is worth your time.
We do not accept payment for placement, and no listing is ranked because it pays a referral. Where an apply link carries a referral parameter we say so on the card.
Never pay to work. No legitimate platform charges a registration, activation, training, security-deposit or “VPN setup” fee, and none pays contributors in crypto or gift cards. No platform needs your bank login, your card details or a scan of your ID sent over WhatsApp or Telegram. Open the company's own domain before you enter anything, and use the link on the card — we do not host applications, and we cannot recover money or an account for you. If a “recruiter” contacts you first with a guaranteed daily payout, treat it as a scam regardless of how official the logo looks.
Worth 60 seconds before you apply: the 14-point scam checklist · how the entry assessments are scored · what an advertised rate is really worth after fees.
Independent, hands-on write-ups about remote AI work — how the roles function, what they realistically pay, and how to avoid the scams that surround this industry.
AI training and data work can be done from home on a flexible schedule, but remote does not mean worldwide. Platforms including Scale AI, Outlier, Mercor, Surge AI, Turing, and DataAnnotation.tech offer remote opportunities, while country and project eligibility vary. Check the current official role and our country guide before applying; DataAnnotation’s reviewed FAQ does not publish a complete country list.
Whether you want a full-time remote career or a flexible side income, these platforms publish openings for every level. All you need to get started is a reliable internet connection, a computer, and strong analytical or writing skills. The work ranges from simple chatbot evaluation and translation projects to advanced RLHF coding and medical data annotation — there is a realistic entry point for most backgrounds, and our beginner's guide walks you through finding yours.
Apply to global AI jobs from anywhere in the world. Platforms hire independent contractors globally, allowing you to work on your own schedule from your home country.
Earn competitive hourly or salary pay. General data annotation pays $15–$40/hr, while credentialed experts (coders, mathematicians, bilingual specialists) can earn $50–$150+/hr.
No prior AI or coding background needed. Entry-level AI training jobs focus on chatbot QA, response rating, and basic labeling, with free training and easy onboarding.
If you code in Python, C++, or Java, or have academic degrees in Math, Law, or Medicine, you can write and review complex code, math proofs, and research to train frontier models.
HowToAIjob, also known as HowToAIjobs.com, is a listings directory for the AI-economy job market — the fast-growing world of AI training, data annotation, model evaluation, and expert-review work that companies like Scale AI, Mercor, and Outlier rely on to build better AI systems. Instead of checking a dozen different company career pages every day, you can scan one page here and see what's currently open, roughly what it pays, and whether it's freelance or full-time.
We don't run any of these jobs ourselves, and we're not affiliated with the companies listed below. We track public hiring pages across the AI training industry and pull the details into one place, updated as listings change. When you click through on a card, you're applying directly with that company, on their own site, under their own terms — we're simply the map, not the destination.
Our team checks public career and hiring pages across dozens of AI companies and platforms for new openings in training, annotation, evaluation, and related freelance or full-time roles.
Each listing gets sorted by type — freelance, full-time, or part-time — tagged with pay range and location where available, and marked "New" when it's been posted recently.
Tap any listing and you'll find a plain-language summary: what the work actually involves, who it tends to suit, and what's worth knowing before you apply — not just a one-line title.
"Visit official page" sends you straight to the company's own application page. We don't collect applications, charge fees, or sit anywhere in the middle of the actual hiring process.
Many AI training and annotation companies run separate hiring tracks, higher pay bands, and region-exclusive projects for people in the United Kingdom and the United States. If you're in the UK or USA, contact us directly and we'll share job details, requirements, and application guidance specific to your region — completely free.
Mention your country (UK / USA) and your background so we can point you to the right openings faster. We never charge for this — it's free guidance, not recruitment.
Yes — every listing links straight to the company's own official careers or hiring page, and we note only public, verifiable openings. We're not affiliated with any of these companies and don't collect applications ourselves, so you always apply directly on their site under their own terms. Always confirm details before starting work or sharing personal information, and never pay anyone to "secure" a job.
Most data annotation, RLHF, and AI training roles are entry-friendly and don't require a technical background — companies like Outlier, DataAnnotation.tech, and Prolific mainly look for strong written English and attention to detail. Some specialist and expert-review roles (medical, legal, coding) do require relevant credentials or domain expertise, which is noted on each listing.
Pay varies widely by platform and specialization — general annotation and evaluation work often advertises $15–$50/hr, while selected current Surge lawyer, journalist, management-consultant and investment-banker role results advertise $200–$400/hr. These are role-specific advertised figures, not typical or guaranteed earnings; each card shows its own captured pay label and scope.
Yes, completely free. We're an independent listings directory, not a recruiter — we don't charge job seekers, and we don't sit between you and the hiring company. Browsing, filtering, and applying through any listing costs nothing.
Yes! Most listings here are 100% remote and open to applicants worldwide, including the United States, India, Canada, United Kingdom, Europe, Latin America, Southeast Asia, and Africa. AI jobs and remote work are inherently borderless. Platforms like Scale AI, Outlier, and Surge AI employ remote AI trainers and data annotators globally. Even if you're in a country with fewer local tech jobs, you can apply for these global AI training positions and work from anywhere.
We check public hiring pages across the AI training industry regularly and update listings as they change — new or recently updated roles are marked "New" on their card, and the "Last updated" date at the top of the page reflects the most recent refresh.
Yes — many AI training companies run separate hiring tracks and higher pay bands for US/UK-based applicants. Check the UK & USA contact section above to reach us directly, and we'll point you to region-specific openings and requirements — free of charge.
Join our Telegram community for daily AI job alerts — completely free.
HowToAIjob collects public opportunities from company career pages and work platforms so job seekers spend less time searching. Availability, location and pay can change; always confirm the current terms on the destination before applying.
Listings are checked on a regular schedule. A review date is not a guarantee that a role remains open between checks.
HowToAIjob does not collect job applications or charge applicants. Apply on the company or platform page linked from each card.
Never pay to secure a job. Verify recruiter domains and contract terms before sharing identity or banking information.
Read the scam checklist →
We use cookies. Essential cookies keep the site working. With your permission we also use analytics and advertising cookies to understand what is read and to support the site. You can change your mind at any time. Privacy policy