
Expert network · Six-month expert fellowships (1099 contract)
Scale AI Human Frontier Collective
About Scale AI Human Frontier Collective
The Human Frontier Collective is Scale AI's program for credentialed experts, separate from its Outlier contributor platform. Fellows in medicine, law, finance, STEM, machine learning and software engineering work on frontier AI research and evaluation projects as fully remote 1099 independent contractors, with the US, UK and Canada as separate cohorts: 9 of its 13 fellow postings give an estimated six months, and the machine learning (US, UK, Canada) and US STEM postings say there is no fixed end date (its Greenhouse board, read 5 Nov 2026). It also hires part-time research advisors (finance, legal, medical), creative advisors who bring their craft to Scale's research teams, part-time subject matter experts who audit quality, and an AI Safety Residency run with Scale Labs Preparedness and Policy (3 months, extendable; posting updated 6 Nov 2026). Its own FAQ puts fellows' rates at $80 to $120 an hour (hfc.scale.com, read 12 Nov 2026).
Platform-stated
, AITraining.jobs lists 10 open Scale AI Human Frontier Collective roles; 3 of them state an hourly rate.
Live Scale AI Human Frontier Collective roles
Updated 13 Nov 2026, 05:07 UTCSimilar roles you can apply to now
The same kind of work on other platforms.

AI Conversation Evaluator (opens the job details)
micro1 · AI Evaluation · Remote· todayDetails
$40–$70/hr
Document Annotation Specialist (PDF) (opens the job details)
SuperAnnotate (SME) · Data Annotation · Remote· 2d agoDetails
up to $10/hr

General AfterQuery Expert (Data Annotator) (opens the job details)
AfterQuery · Data Annotation · Remote· 1y+ agoDetails
$43/hr
We may earn a referral fee when you join through these links. It never changes your pay.
Subject Matter Expert (opens the job details)
Domain Experts · Canada; United States· 5mo agoDetails
- $100/hr
Creative Advisor - Human Frontier Collective (opens the job details)
Domain Experts · United States· 2w agoDetails
$300/hrResearch Advisor - Human Frontier Collective (US) (opens the job details)
Studies & Surveys · United States· 6mo agoDetails
$300/hr
Some links on AITraining.jobs are referral links. If you sign up through one, AITraining.jobs may earn a commission at no extra cost to you. It never changes what a platform pays you.
Scale AI Human Frontier Collective jobs by country
Its open roles that accept applicants in each country.
Get new Scale AI Human Frontier Collective roles by email
We check the Scale AI Human Frontier Collective job board every hour and email you its newest roles each week, plus similar paid roles elsewhere. Free, unsubscribe anytime.
Is Scale AI Human Frontier Collective legit?
Researched Nov 2026 · 23 sourcesYes. The Human Frontier Collective is Scale AI's real, paid fellowship for credentialed experts, and its FAQ puts hourly rates at $80 to $120, but fellows are 1099 contractors paid per project under Outlier's terms. No fellow posting promises hours: fellows are matched to projects one at a time, each with its own rate, and the terms of HFC's platform let Scale delay pay in an investigation, withhold it for a violation, or close accounts without notice. It pays $80–$120/hr (its FAQ); research and creative advisors $300/hr; AI Safety Residency $100/hr (platform-stated).
Actively listing 10 open roles todayHFC is Scale AI's expert fellowship, live since at least March 2025, when its site called it a part-time program paying $100 an hour. Outlier's Terms of Use, with Scale AI, Inc. and Smart Ecosystem Inc., govern its platform. Meta paid $14.3 billion for 49% of Scale in June 2025, valuing it at about $29 billion; founder Alexandr Wang left for Meta but stays chairman, and Francis deSouza became CEO on 10 August 2026.
HFC's FAQ: hourly rates of $80 to $120 by field and scope, more as bonuses for success; advisors are offered $300 an hour, AI safety residents $100. Work runs on HFC's platform, Outlier's (paid weekly on Tuesdays by PayPal, Airtm or ACH) or a partner lab's. Fellows are 1099 contractors; HFC publishes no payout schedule.
No worker reviews of the Human Frontier Collective found (2026-11-12)
10 open roles on Scale AI Human Frontier Collective, independently checked 13 Nov 2026.
The United States, the United Kingdom and Canada, each fellowship posted for its country; US roles require US work authorization, and the AI Safety Residency is remote within the US although its board entry also lists Canada and the UK (read 5 Nov 2026). Outlier's Terms of Use (effective 11 May 2026), which name HFC's platform, also rule out Ukraine, Cuba, Iran, North Korea, Sudan and Syria (read 5 Nov 2026).
No HFC-specific impersonation cases turned up. HFC says it does no marketing or email campaigns, its roles open on Scale's Greenhouse board, and its support address is hfcsupport@scale.com. The one review on an unclaimed Trustpilot page for scale.com (November 2025) describes a home job in Scale's name that kept deposits. The FTC: never pay to get a job.
What people praise
- HFC's own FAQ states hourly rates of $80 to $120 by field and scope, and says far greater pay or bonuses are available contingent on success (read 12 Nov 2026).
- Fellows co-author published work with Scale's researchers: HFC lists Humanity's Last Exam (in Nature, January 2026), the Professional Reasoning Benchmark and SciPredict among its results.
- The Machine Learning and US STEM postings set no minimum commitment: you decide whether to take each project, and most fellows spend 10 to 25 hours a week on their own schedule.
What people complain about
- No fellow posting states a rate: pay 'varies across platforms' by project, scope, skillset and location, and the Machine Learning posting says you learn it when matched to a project.
- Outlier's Terms of Use, which govern HFC's platform, let Scale delay pay during an investigation, withhold some or all of it for a violation in its 'sole determination', and close accounts without notice.
- In the US, disputes go to binding individual arbitration with no class actions, unless you email an opt-out within 30 days of accepting the terms (Outlier's Terms of Use, effective 11 May 2026).
- If you are turned down, Scale waits 90 days before it reconsiders you for the same role.
What has gone wrong
- October 2026. On 27 October 2026 the privacy group noyb filed five GDPR complaints against Scale AI in Austria, France, Italy and the Netherlands, saying Outlier, whose platform also hosts HFC projects, tracks workers' devices, requires repeated face-video ID checks and ignored data requests. Scale called the claims 'unfounded and inaccurate'.
- December 2024. A class action filed in December 2024, McKinney v. Scale AI, says contributors were misclassified as contractors. A $12.5 million settlement covers people who worked for Scale, Smart Ecosystem or through HireArt from California (10 Dec 2020 to 28 Feb 2026); the court granted final approval on 30 Oct 2026, though its site's home page still says it has not decided (13 Nov 2026).
- July 2025. After Meta's June 2025 investment, several of Scale's largest data customers cut ties with it. In July 2025 Scale laid off about 200 staff, around 14%, and ended work with 500 contractors, saying it had scaled its data labeling business too quickly.
- June 2025. Business Insider reported in June 2025 that Scale kept work for clients such as Meta and xAI in public Google Docs, some marked confidential, leaving thousands of contractors' names, private email addresses and performance notes open. Scale launched an investigation and locked thousands of files; BI saw no sign of a breach.
Best for: PhDs, board-certified doctors, lawyers, finance professionals, ML researchers and senior engineers in the US, UK or Canada who want remote, self-scheduled AI projects and a chance to co-author research. Not for: Anyone who needs a set rate before applying, a salary, benefits or steady hours, needs a US visa (US fellowships sponsor none), or lives outside the US, UK and Canada, where no fellowship is posted.
Honest catches
- Pay here is platform-stated, not measured live from a job feed, so treat it as a benchmark rather than a guaranteed rate.
Sources for this review
- Human Frontier Collective: home page and FAQ ($80 to $120 an hour, projects on HFC's platform, Outlier's or a partner lab's, applicant loops within 2 weeks, 92% doctorates; read 12 Nov 2026)
- Human Frontier Collective: Explore the Collective (no marketing or email campaigns, open positions; read 12 Nov 2026)
- Human Frontier Collective: Research that Ships (published benchmarks; read 12 Nov 2026)
- Human Frontier Collective home page, archived 8 Mar 2025 (part-time program for PhDs, $100 an hour)
- Scale AI job board on Greenhouse (18 Human Frontier Collective and Specialist postings; read 12 Nov 2026)
- Scale AI: Machine Learning Fellow, Human Frontier Collective (US; no fixed end date, pay with each project match; read 12 Nov 2026)
- Scale AI: Medical Fellow, Human Frontier Collective (US; six months, challenge and interview, 90-day wait; read 12 Nov 2026)
- Scale AI: Research Advisor, Human Frontier Collective (US; $300 an hour; read 12 Nov 2026)
- Scale AI: AI Safety Residency, SLPreP (Human Frontier Collective; $100 an hour, 20 hours a week for 3 months; updated 6 Nov 2026)
- Outlier: Terms of Use (effective 11 May 2026; covers platform.hfc.scale.com, HFC support hfcsupport@scale.com; read 12 Nov 2026)
- Outlier: FAQ (weekly Tuesday payouts by PayPal, AirTM or ACH; read 12 Nov 2026)
- Scale AI: Scale AI announces next phase of company's evolution (12 Jun 2025)
- Reuters via Hargreaves Lansdown: Meta poaches Scale AI CEO after taking multibillion-dollar stake (13 Jun 2025)
- Scale AI: Scale AI appoints Francis deSouza as CEO (30 Jul 2026; Wang stays chairman)
- TechCrunch: Scale AI lays off 14% of staff, largely in data-labeling business (16 Jul 2025)
- Business Insider (archived): BI revealed that Scale AI exposed sensitive data about Big Tech clients (26 Jun 2025)
- King & Siegel LLP: Scale AI pays $12.5 million over misclassification claims
- McKinney v. Scale AI settlement website (its home page still says the court has not decided; read 13 Nov 2026)
- noyb: AI training platform Outlier illegally surveils European platform workers (27 Oct 2026)
- The Register: EU privacy advocates accuse US data annotation firm of snooping on contractors (27 Oct 2026; updated with Scale's response)
- Trustpilot: System Scale (scale.com), unclaimed profile with one review (read 12 Nov 2026)
- FTC Consumer Alert: That job offer text is probably a scam (30 Apr 2026)
- Superior Court of California, San Francisco: Order Granting Final Approval of Class Action Settlement, McKinney v. Scale AI, CGC-24-620481 (filed 30 Oct 2026, posted by the settlement administrator; read 13 Nov 2026)
What Scale AI Human Frontier Collective pays
Platform-stated · as of 2026-11-12Researched rate card, platform-stated, as of 2026-11-12
From Scale AI's Greenhouse board (job-boards.greenhouse.io/scaleai, its Human Frontier Collective and Specialist departments, 18 postings read 5 Nov 2026): the US and UK research advisor postings say 'We are offering $300/hr' for part-time contract work, the Creative Advisor posting (US, first published 23 Oct 2026) says 'Paid $300 hourly', with time and scope set at interview, and the Safety Research Scholar posting (first published 16 Oct 2026, remote within the US, 20 hours a week for 3 months) says 'Compensation: $100/hour'. The 13 fellow postings (medical, legal, finance, STEM, machine learning, software engineering) state no figure: rates 'vary across platforms' by project, scope, skillset and location, and the ML and US STEM fellows are told the pay with each project match. The Subject Matter Expert posting (US and Canada, part-time hourly) states no rate either, and none of the 18 carries a pay range in the board's pay-transparency data. Accel's portfolio job board, which copies Scale's postings, shows USD 80 an hour on the US Machine Learning Fellow (posted 13 Feb 2026) and USD 80 to 120 on the US STEM Fellow (posted 7 Jun 2025) (jobs.accel.com, read 5 Nov 2026); Scale's own postings do not state these figures.
3 of its 10 open roles state a rate, too few to call any of those rates typical of Scale AI Human Frontier Collective, so the figure above is the researched one. What each kind of work states:
| Kind of work | Open roles | Rates stated | Highest rate |
|---|---|---|---|
| Domain Experts | 5 | $300/hr1 of 5 state a rate | $300/hr |
| AI Evaluation | 1 | No rate stated | None stated |
| Coding | 1 | No rate stated | None stated |
| Math & STEM | 1 | No rate stated | None stated |
| Red-Teaming | 1 | $100/hr1 rate | $100/hr |
| Studies & Surveys | 1 | $300/hr1 rate | $300/hr |
| Role | Kind of work | Researched pay |
|---|---|---|
| Research Advisor (US, UK) | Domain Experts | $300/hr |
| Creative Advisor (US) | Domain Experts | $300/hr |
| Medical, Legal or Finance Fellow | Domain Experts | Rate by project |
| STEM Fellow (US, UK) | Math & STEM | Rate by project |
| Machine Learning or SWE Fellow | Coding | Rate by project |
| Subject Matter Expert (quality audits) | AI Evaluation | Hourly, rate not stated |
| Safety Research Scholar (US, AI safety research) | Red-Teaming | $100/hr |
The work Scale AI Human Frontier Collective hires for
Each links to what the work involves, who hires for it, and what it pays across the market.
New to this? Start with the learning guides or compare pay across every platform.
How long Scale AI Human Frontier Collective keeps roles listed
Scale AI Human Frontier Collective's row in How Long AI Training Roles Stay Open, our research over 3 Aug 2026 to 12 Nov 2026, rebuilt every day. Every figure opens the roles it counts, and each table its place among the others in the report.
How long roles stay listed, by platform
On which platforms do AI training roles come down fastest?
| Platform | Still listed after 2 weeks | After 4 weeks | Median lifetime (days) |
|---|---|---|---|
| All platforms | 61.0% | 25.0 | |
| Scale AI Human Frontier Collective | 100.0% | n/a | n/a |
How to get started on Scale AI Human Frontier Collective
Researched Nov 2026Apply to the fellowship for your field and country through hfc.scale.com, which opens Scale AI's Greenhouse posting; Scale reviews your background, emails selected applicants an assessment and a screening interview, says it tends to finish within 2 weeks, and invites those who pass, who are then matched to paid projects one at a time.
Who can join
- Medical Fellow: an MD or DO with board certification. Residents, practising clinicians and people with extensive clinical experience are welcome, in internal medicine, emergency medicine, radiology, surgery, psychiatry and related fields; no AI experience is needed.
- Legal Fellow: a JD or LLM, or a current 2L or 3L place at what the posting calls a prestigious program. Bar admission in any jurisdiction is strongly preferred, and the posting names attorneys, clerks, counsel, legal researchers and law professors.
- Finance Fellow: a PhD, master's or BBA in finance, economics or a related field and 5+ years in FP&A, investment banking, portfolio management, financial consulting or equivalent research; a CFA or CPA is strongly preferred. On 12 Nov 2026 the board carried only the UK posting.
- STEM Fellow: PhDs, postdoctoral researchers and professors with expertise in fields such as computer science, machine learning, physics, biology, chemistry, mathematics or the humanities.
- Machine Learning Fellow: a PhD or postdoctoral research in machine learning or computer science, or senior industry experience in machine learning, with strong Python and PyTorch or TensorFlow (AWS, Docker and LangChain are a plus).
- SWE Fellow: 5+ years of software engineering, or a PhD, postdoc or faculty post in computer science, with deep expertise in Python, Java, Rust, Swift, JavaScript or C++.
- For a US fellowship, authorization to work in the United States: Scale sponsors no visas for these roles.
- For a UK or Canada fellowship, authorization to work in the country you live in.
- Be at least 18, or the age of majority where you live, and hold one account only. Outlier's Terms of Use, which govern HFC's platform, make you an independent contractor and ask for valid payment details in your own name.
- Be ready to prove where you are. During onboarding you may have to verify your identity and location with a government-issued ID; that place becomes your Primary Country, the only one you may work from.
Step by step
- 1
Pick your track and country
HFC's site lists tracks for research advisors, medicine, math, computer science and ML, physics, chemistry, law, biology and creative advisors, and its Apply buttons open the matching posting on Scale AI's Greenhouse board. Apply to the posting for the country you live in.
- 2
Apply on the posting
Scale reviews applications on a rolling basis. Choose carefully: if you are turned down, Scale waits 90 days before it reconsiders you for the same role.
- 3
Wait for the background review
HFC says it personally reviews your background and how well you fit the program, then sends the next step by email: an assessment and a screening interview.
- 4
Complete the challenge
On the medical, legal, finance and UK STEM postings, selected candidates do a brief task designed to reflect HFC work and evaluate their domain expertise. The ML, SWE and US STEM postings list only the interview.
- 5
Interview
You discuss your research experience, your professional background and your fit with Scale's mission to advance human-centered AI. On the ML and US STEM postings, if Scale invited you directly, your invitation says whether this step applies.
- 6
Accept the invitation
Successful candidates are invited to join the Human Frontier Collective Fellowship, which also opens its newsletters, discussion channels, virtual sessions and in-person events.
- 7
Get matched to projects
Scale matches you to projects from its partners, from evaluating AI models to designing experiments, building RL environments and co-authoring papers. Each match notification states the pay, there is no minimum commitment, and you decide whether to take a project.
- 8
Work where the project is hosted
HFC's FAQ says projects run on HFC's own platform, on Outlier's platform or in a partner lab's environment, and that expectations and time commitment are set before you say yes.
A brief domain challenge (on most tracks) and a screening interview
HFC's FAQ says the step after the review of your application arrives by email and involves an assessment and a screening interview. The medical, legal, finance and UK STEM postings describe the assessment as a brief task designed to reflect HFC work and evaluate your domain expertise; the ML, SWE and US STEM postings list only the interview, about your research experience, professional background and fit with Scale's mission. Research advisors meet Scale's research team, and AI safety residents do an ML research take-home, a 30-minute research discussion and a 15-minute hiring manager chat.
Format
Online, by email invitation: a brief task in your field, then a live interview. Creative Advisors instead share one past project for a free portfolio review, then have a paid conversation with Scale's research and ML team.
How to prepare
- Apply to the track that matches your credential: each posting names its own bar, such as board certification for medicine, a JD or LLM for law, and 5+ years of practice for finance and software engineering.
- Expect the challenge to look like the work: the medical posting describes designing complex clinical scenarios and evaluating model decisions for safety and accuracy, the legal posting evaluating model outputs for accuracy and regulatory alignment.
- Prepare to talk about your own research or casework in detail, since the interview covers your research experience and professional background.
- AI safety residents: Scale says the take-home is judged more on your reasoning and how you interpret results than on polished or conclusive outcomes.
- Creative Advisors: have a portfolio link and one past project you are proud of ready; nothing is used without a signed agreement and payment.
Why people get rejected
- Scale does not publish why applicants are turned down; the bar it states is each posting's 'Who Should Apply' list (for medicine, an MD or DO with board certification) and, for US fellowships, US work authorization with no visa sponsorship.
HFC's FAQ says it tends to complete all applicant loops within 2 weeks of submission: a personal review of your background, then an emailed assessment and screening interview. It publishes no wait between joining and a first project.
HFC's FAQ says pay depends on your field and the scope of work: hourly rates vary from $80 to $120, and far greater pay or bonuses are available contingent on success. Projects are hosted on HFC's platform, Outlier's platform or a partner lab's environment, and HFC publishes no payout schedule or method of its own. How often other platforms pay
Tips to get accepted
- Apply through hfc.scale.com or Scale's Greenhouse board. HFC says it does no marketing or email campaigns and grows mostly by word of mouth, with fellows referring peers through its open roles.
- Read each project's rate and platform before you accept: the match notification states the pay, and the postings say rates vary across platforms by project, scope, skillset and location.
- If a project runs on Outlier, Outlier's FAQ says it pays weekly on Tuesdays for work from the previous Tuesday to Monday (midnight UTC), by PayPal, AirTM or ACH bank transfer.
- Set money aside for tax: the terms say Scale withholds no income tax or social contributions, and that you must register as self-employed where your law requires it.
- In the US, disputes go to individual arbitration unless you opt out within 30 days of accepting the terms, by emailing legalnotices@scale.com from your account's address with your full name and city.
- Work only from your verified Primary Country: if you work elsewhere or move without updating it, your account may be paused until the new location is verified.
- HFC's support address is hfcsupport@scale.com; Outlier's support chatbot and Remotasks' address serve those platforms.
What isn't public
- HFC publishes no payout schedule, payment method or minimum for work on its own platform or in a partner lab's environment; only Outlier's weekly schedule is public, and HFC's member platform (platform.hfc.scale.com) needs a sign-in we did not make.
- None of the 13 fellow postings on Scale's board on 12 Nov 2026 states a rate. The $80 to $120 range is in HFC's FAQ only, and it does not say which fields or projects pay at either end, or what the success-based bonuses are.
- Scale publishes no acceptance rate, no count of fellows and no figure for how many projects or paid hours a typical fellow gets; the fellow postings set no minimum and promise none.
- The challenge's length, format and scoring are not published, and the pages differ: HFC's FAQ says applicants get an assessment and a screening interview, while the ML, SWE and US STEM postings list only an interview.
- Postings differ on length: 9 of the 13 fellow postings give an estimated six months with possible extension, while the Machine Learning (US, UK, Canada) and US STEM postings say there is no fixed end date.
- The legal pages linked from hfc.scale.com are older Outlier versions (its Terms of Use and Community Guidelines dated 15 May 2023), while Outlier's current Terms of Use, effective 11 May 2026, name platform.hfc.scale.com; which version you accept is shown only at sign-up.
- HFC's Collective page lists a Finance Fellow for the US and the UK, but only the UK posting was on Scale's Greenhouse board on 12 Nov 2026.
- We found no independent reviews by HFC fellows on Trustpilot, Glassdoor or Indeed and no press coverage of the program itself, and could not read Reddit or Glassdoor directly; the testimonials on HFC's site are chosen by Scale.
Sources (17)
- Human Frontier Collective: home page and FAQ ($80 to $120 an hour, projects on HFC's platform, Outlier's or a partner lab's, applicant loops within 2 weeks; read 12 Nov 2026)
- Human Frontier Collective: Become a Fellow (tracks by field; read 12 Nov 2026)
- Human Frontier Collective: Explore the Collective (no marketing or email campaigns; read 12 Nov 2026)
- Human Frontier Collective: Working Location Policy (7 Feb 2024; read 12 Nov 2026)
- Scale AI job board on Greenhouse (13 fellow postings across the US, the UK and Canada; read 12 Nov 2026)
- Scale AI: Medical Fellow, Human Frontier Collective (US; read 12 Nov 2026)
- Scale AI: Legal Fellow, Human Frontier Collective (US; read 12 Nov 2026)
- Scale AI: Finance Fellow, Human Frontier Collective (UK; read 12 Nov 2026)
- Scale AI: STEM Fellow, Human Frontier Collective (US; updated 6 Nov 2026, read 12 Nov 2026)
- Scale AI: Machine Learning Fellow, Human Frontier Collective (US; read 12 Nov 2026)
- Scale AI: SWE Fellow, Human Frontier Collective (US; read 12 Nov 2026)
- Scale AI: SWE Fellow, Human Frontier Collective (UK; read 12 Nov 2026)
- Scale AI: Research Advisor, Human Frontier Collective (US; $300 an hour; read 12 Nov 2026)
- Scale AI: Creative Advisor, Human Frontier Collective (US; $300 an hour; read 12 Nov 2026)
- Scale AI: AI Safety Residency, SLPreP (Human Frontier Collective; $100 an hour; updated 6 Nov 2026)
- Outlier: Terms of Use (effective 11 May 2026; covers platform.hfc.scale.com; read 12 Nov 2026)
- Outlier: FAQ (weekly Tuesday payouts by PayPal, AirTM or ACH; read 12 Nov 2026)
From the forum
Ask the people doing Scale AI Human Frontier Collective work what only they can tell you.
Frequently asked questions
Is Scale AI Human Frontier Collective legit?
Yes. The Human Frontier Collective is Scale AI's real, paid fellowship for credentialed experts, and its FAQ puts hourly rates at $80 to $120, but fellows are 1099 contractors paid per project under Outlier's terms. No fellow posting promises hours: fellows are matched to projects one at a time, each with its own rate, and the terms of HFC's platform let Scale delay pay in an investigation, withhold it for a violation, or close accounts without notice.
Is Scale AI Human Frontier Collective real or a scam?
Real. Scale AI Human Frontier Collective is a real company, not a scam. No HFC-specific impersonation cases turned up. HFC says it does no marketing or email campaigns, its roles open on Scale's Greenhouse board, and its support address is hfcsupport@scale.com. The one review on an unclaimed Trustpilot page for scale.com (November 2025) describes a home job in Scale's name that kept deposits. The FTC: never pay to get a job.
Is Scale AI Human Frontier Collective hiring right now?
Yes. As of 13 Nov 2026, AITraining.jobs lists 10 open Scale AI Human Frontier Collective roles; 3 of them state an hourly rate. Its job board is read every hour, so the count moves as roles open and close.
Does Scale AI Human Frontier Collective pay?
Yes. Run by Scale AI, which also runs Outlier. The fellowships are newer than Outlier and have no independent worker reviews of their own; Outlier's reviews are on its own page. Payouts: 1099 independent contract, paid per project: work runs on HFC's platform, on Outlier's (weekly Tuesday payouts by PayPal, AirTM or ACH) or in a partner lab's environment; HFC publishes no payout schedule of its own (its FAQ and Outlier's terms, read 12 Nov 2026). HFC's FAQ says pay depends on your field and the scope of work: hourly rates vary from $80 to $120, and far greater pay or bonuses are available contingent on success. Projects are hosted on HFC's platform, Outlier's platform or a partner lab's environment, and HFC publishes no payout schedule or method of its own.
How much does Scale AI Human Frontier Collective pay?
$80–$120/hr (its FAQ); research and creative advisors $300/hr; AI Safety Residency $100/hr. From Scale AI's Greenhouse board (job-boards.greenhouse.io/scaleai, its Human Frontier Collective and Specialist departments, 18 postings read 5 Nov 2026): the US and UK research advisor postings say 'We are offering $300/hr' for part-time contract work, the Creative Advisor posting (US, first published 23 Oct 2026) says 'Paid $300 hourly', with time and scope set at interview, and the Safety Research Scholar posting (first published 16 Oct 2026, remote within the US, 20 hours a week for 3 months) says 'Compensation: $100/hour'. The 13 fellow postings (medical, legal, finance, STEM, machine learning, software engineering) state no figure: rates 'vary across platforms' by project, scope, skillset and location, and the ML and US STEM fellows are told the pay with each project match. The Subject Matter Expert posting (US and Canada, part-time hourly) states no rate either, and none of the 18 carries a pay range in the board's pay-transparency data. Accel's portfolio job board, which copies Scale's postings, shows USD 80 an hour on the US Machine Learning Fellow (posted 13 Feb 2026) and USD 80 to 120 on the US STEM Fellow (posted 7 Jun 2025) (jobs.accel.com, read 5 Nov 2026); Scale's own postings do not state these figures.
Is Scale AI Human Frontier Collective worth it?
Scale AI Human Frontier Collective pays $80–$120/hr (its FAQ); research and creative advisors $300/hr; AI Safety Residency $100/hr (platform-stated). It rewards credentialed experts who clear its assessment far more than open generalist work. Worker sentiment: No worker reviews of the Human Frontier Collective found (2026-11-12).
How does Scale AI Human Frontier Collective pay out?
1099 independent contract, paid per project: work runs on HFC's platform, on Outlier's (weekly Tuesday payouts by PayPal, AirTM or ACH) or in a partner lab's environment; HFC publishes no payout schedule of its own (its FAQ and Outlier's terms, read 12 Nov 2026).
How often does Scale AI Human Frontier Collective pay?
Scale AI Human Frontier Collective publishes no pay schedule, so ask for it in writing before you start. How often other platforms pay.
Does Scale AI Human Frontier Collective pay through PayPal?
Scale AI Human Frontier Collective does not publish its payout method. How other platforms pay out.
Is Scale AI Human Frontier Collective 1099 or W-2, and what tax form does it send?
Contractor (1099) work: Scale AI Human Frontier Collective engages independent contractors, not employees, and you handle your own taxes. Its fellows are 1099 contractors.
Where can you work for Scale AI Human Frontier Collective?
The United States, the United Kingdom and Canada, each fellowship posted for its country; US roles require US work authorization, and the AI Safety Residency is remote within the US although its board entry also lists Canada and the UK (read 5 Nov 2026). Outlier's Terms of Use (effective 11 May 2026), which name HFC's platform, also rule out Ukraine, Cuba, Iran, North Korea, Sudan and Syria (read 5 Nov 2026).
How do I join Scale AI's Human Frontier Collective?
Apply to the fellowship for your field and country through hfc.scale.com, which opens the posting on Scale AI's Greenhouse board. Scale reviews your background, emails selected applicants an assessment and a screening interview, and invites those who pass to join; it says it tends to complete all applicant loops within 2 weeks.
How much does the Human Frontier Collective pay?
HFC's FAQ says hourly rates vary from $80 to $120 depending on your field and the scope of work, with far greater pay or bonuses available contingent on success. The fellow postings themselves state no rate; each project's pay comes with its match notification.
What do HFC's Research Advisors earn?
$300 an hour across finance, legal and medical, for part-time contract work of about 5 to 20 hours a week. The posting looks for senior people such as bulge-bracket or elite-boutique finance leaders, V100 partners and board-certified physicians with 5+ years after residency, and its interview is with Scale's research team.
What is HFC's Creative Advisor role?
A paid, part-time US role for leading creators who bring their craft to Scale's research and ML teams: commissioned briefs, reviews of model outputs and defining what good work looks like in their field. It pays $300 an hour, with pay, time and scope shared at interview; you apply with your background and a portfolio link.
How do HFC fellows get paid?
Per project, as independent contractors. HFC's FAQ says projects are hosted on HFC's own platform, on Outlier's platform or in a partner lab's environment; HFC publishes no payout schedule or method for its own work.
Is the Human Frontier Collective part of Outlier?
It is a separate Scale AI program for experts, with its own site and support address (hfcsupport@scale.com), but one Terms of Use, with Scale AI, Inc. and Smart Ecosystem Inc. (which does business as Outlier AI), governs outlier.ai, remotasks.com and platform.hfc.scale.com.
Does the Human Frontier Collective have an assessment?
Usually. The medical, legal, finance and UK STEM postings ask selected candidates for a brief task that reflects HFC work and tests their domain expertise, followed by an interview; the ML, SWE and US STEM postings list only the interview.
Do I need a PhD to join HFC?
Not on every track. HFC's FAQ says it is for researchers, industry leaders and senior practitioners with deep expertise in a field, and its site says 92% of fellows hold or are pursuing a PhD, MD, JD or other terminal degree. Each posting names its own requirements.
Can I join HFC from outside the US?
From the UK or Canada, yes. On 12 Nov 2026 Scale's board listed UK fellowships for finance, law, medicine, STEM, machine learning and software engineering, and Canadian ones for machine learning and software engineering, but no fellowship for any other country.
How many hours do HFC fellows work?
You choose. The Machine Learning posting sets no minimum commitment, lets you decide whether to take each project, and says most fellows spend 10 to 25 hours a week on a schedule they set themselves.
Is an HFC fellowship a job or a contract?
A contract. The terms that govern HFC's platform make you an independent contractor with no benefits or perks of employment, and Scale withholds no income tax or social contributions from your pay.
What is HFC's AI Safety Residency?
A Human Frontier Collective program with Scale Labs Preparedness and Policy (SLPreP) for PhD-level researchers: mentor-guided AI safety projects, usually aimed at a conference paper, 20 hours a week for 3 months (extendable to 6 to 12 months for exceptional candidates), remote within the US, at $100 an hour.
Compare Scale AI Human Frontier Collective with
Expert AI-training contracts for degreed professionals: STEM, coding, CAD, medicine, law, finance.
$60–$150/hrTypical pay, livePremium RLHF and expert data for frontier labs, by selective application.
$60–$400/hrTypical pay, liveExpert network for frontier AI labs (OpenAI's logo on its home page) where experts set their own hourly rate for RLHF and evals.
$55–$125/hrTypical pay, liveOrdered by what applying is worth, lifted by a referral boost for the platforms that pay us a referral fee; it never changes what a platform pays you.
Short of experts at Scale AI Human Frontier Collective? We find the experts (you contract them yourself) or we run the project (we deliver the expert work).