Ethos

Quick facts

Company

Ethos

Service type

Data annotation / AI training data

Specialties

text, legal, healthcare, finance, software eng. (code / dev tasks), stem

Headquarters

Worldwide

Hiring status

Both: hires workers and takes vendor projects

Website

Website

Careers

Careers page

Profile last verified

2026-09-16

Application process overview

Ethos is not a classic annotation marketplace. It is an expert network: companies, AI labs and investment firms come to Ethos looking for people with specific professional experience, and Ethos matches them from its pool. The work ranges from AI model training and evaluation, such as writing realistic workplace scenarios and grading model outputs against professional standards, to paid expert calls and advisory projects. There is **no application form and no CV upload**. You sign up, then take a conversational voice interview with Ethos's AI agent, which asks about your career history, the specific work you have done, the tools and regulations you know and how many hours you can realistically give. That transcript becomes the profile Ethos's matching engine searches. From there you wait to be matched to an opportunity, and some projects add their own screening step before you start.

Key findings

* **Voice interview instead of a form.** Plan for about 20 to 30 minutes with the AI agent. Be specific: clients search for precise experience such as "IP litigation", not "legal work", so vague answers produce weak matches. * **Signing up is not a job.** Ethos's own terms make clear that joining does not guarantee work, minimum hours or income. Reviewers report waits of one week to a month between the interview and any next step, and some never receive a match. * **Pay is high but varies widely by domain.** Tracked listings run from about $35 to $225 an hour, with a median near $80. Healthcare, compliance and research roles often sit around $100 an hour, and selective legal roles reach $225. Treat an advertised rate and weekly hour cap as a ceiling, not a promise. * **Hours are reviewed before they are paid.** Payment goes by bank transfer to an account in your own name, on a cycle set per engagement or at project completion. Contributors describe monthly payouts, and say logged hours can sit in quality review before release. * **Experienced professionals only.** Many listings ask for four or more years of experience, and specialist legal, clinical and management projects often want five to nine. Prior AI training experience is not required. * **Open worldwide, with checks.** You must be 18 or over and pass identity verification and sanctions screening. Contributors outside the US report being accepted, but Ethos decides case by case which countries it can pay. * **Strict confidentiality.** You work as an independent contractor, and sharing material non-public information or an employer's confidential material is grounds for removal.

Conclusion

Ethos is a well-funded, London-based company. It was founded in 2024 by James Lo, formerly of McKinsey and SoftBank, and Daniel Mankowitz, a former DeepMind researcher. It raised a $3.25M seed round led by General Catalyst and a $22.75M Series A led by a16z in May 2026. Its Trustpilot score of 4.2 comes with many reports of payments arriving on time, and the team is visibly active on its own subreddit. The main risk is time rather than money. Ethos was reported to be onboarding around 35,000 people a week in 2026, so matching is competitive and task queues can be full or empty. A recurring complaint is that the long AI interview feels like unpaid work when no project follows. It is worth the interview if you have deep, specific professional experience, but treat Ethos as one source of well-paid project work rather than a steady income.