Answer · AI training work

Is DataAnnotation legit? Pay, reviews and how to get accepted (2026)

DataAnnotation pays people to write, rate and fact-check AI chatbot answers. It's a real business that pays real money, but the details matter: one shot at the entry test, no guaranteed work, and a pending lawsuit over how its contractors are classified.

Contents

The evidence that it's real

  • It pays, and says how much. DataAnnotation's FAQ says: "We have paid hundreds of millions of dollars to contractors since 2020." That is the company's own claim; it doesn't publish audited figures. It also says it connects "over 1 million contractors globally" with AI companies.
  • It never charges you. "We will never ask for money from you for anything." Payments go to PayPal, and deposits arrive "within a few days after you request them."
  • There's a real company behind it. A 2025 court filing names the operator as Surge Labs, Inc. dba DataAnnotation (complaint, San Francisco Superior Court). The complaint describes Surge AI's business as training AI systems for client companies such as OpenAI and Meta.
  • Workers rate it reasonably well. On 20 Oct 2026 its Trustpilot page showed a 3.9 rating from 1,975 reviews Self-reported. Trustpilot reviews are self-reported, and the platform says it doesn't fact-check them.

What it pays

DataAnnotation publishes starting rates by track. From its FAQ, read 20 Oct 2026 Verified:

Track Starting rate Who qualifies
General $25–$50/hr Bachelor's degree or equivalent experience; strong writing and reasoning
Multilingual $25–$40/hr Native fluency in another language besides English
Coding $40–$150+/hr Programming experience (Python, JavaScript, etc.); LeetCode-style problems
STEM $40–$125+/hr Master's/PhD, or bachelor's + 10 years' experience
Professional $40–$125+/hr Licensed credentials in law, finance or medicine

Its homepage lists 112 open roles, with the generalist role at "up to $50/hr" and expert roles "up to $75–$125+/hr" (homepage, 20 Oct 2026). On 17 Oct the FAQ listed multilingual work at "$20+"; by 20 Oct it said $25–$40. Rates on the site change, so trust the rate shown on each project in your dashboard.

What that can add up to Estimate: at $25/hour for 6 hours a week, about $650 a month (25 × 6 × 4.33). At $40/hour for 10 hours a week, about $1,730. Both drop to $0 in any week with no projects. These are our calculations, not DataAnnotation figures.

The milestone posts on its homepage ("Just hit 75k!!", "Hit the 10k mark in one month") are testimonials the company chose to publish Self-reported. Treat them as best cases.

What to watch out for

  • One attempt at the test. "You can only take the Starter Assessment once. There are no retakes or second chances." It takes about an hour; specialist tests take one to two.
  • No guaranteed work. The FAQ says work availability "depends on both the demand for your skills and your past performance," and "subpar or problematic contributions may result in loss of access to future projects."
  • Little support for applicants. "Given the number of applications to DataAnnotation, we are unable to provide direct responses for every applicant." If you haven't heard back, the FAQ says your application is still under review or wasn't accepted.
  • ID verification. You'll verify a physical government photo ID and a selfie through Persona, with any VPN switched off.
  • A pending class action. Cavalier v. Surge Labs, Inc. (San Francisco Superior Court, CGC-25-625502), filed 20 May 2025, alleges data annotators are misclassified as independent contractors and claims unpaid minimum wage, including for training time (complaint). These are allegations, not findings. We found no ruling or settlement in public sources as of 20 Oct 2026. Scale AI, which runs Outlier, has a separate $12.5 million classification settlement awaiting a final approval hearing on 30 Oct 2026 (settlement site).

How to spot a fake "DataAnnotation" offer

Scammers copy the name. The FTC warns about "task scams" that pay small amounts at first and then ask you to deposit money to unlock earnings (FTC). Real DataAnnotation work:

  • Starts with an application on dataannotation.tech, not a WhatsApp, Telegram or text message from a "recruiter".
  • Never asks you to pay, buy training or deposit crypto.
  • Pays to PayPal under the email you registered with.

How to get accepted

  1. Pick the track you truly qualify for. You choose a Starter Assessment by track; coding, STEM and professional tracks pay more but test harder.
  2. Take your time. The FAQ says reviewers check "thoroughness and accuracy rather than speed," and successful applicants "spend more time crafting thorough answers."
  3. Write like an editor. Explain why an AI answer is wrong, with specifics. Fact-check claims instead of guessing.
  4. Read every instruction twice. You can't retake it.
  5. Prepare your ID and PayPal with the same email, so approval isn't held up.
  6. Apply to two or three platforms, because any one can go quiet. Compare DataAnnotation vs Outlier vs Alignerr, or see how much AI training jobs pay.

We have no affiliate or referral relationship with DataAnnotation. The link to DataAnnotation goes to its own site.

FAQ

Is DataAnnotation a scam?

No. It's run by Surge Labs, charges no fees, pays through PayPal and says it has paid contractors hundreds of millions of dollars since 2020. Fake recruiters do use its name, so apply only at dataannotation.tech and never pay anything.

How much does DataAnnotation pay per hour?

Its FAQ lists starting rates of $25–$50/hour for general work, $25–$40 for multilingual, $40–$150+ for coding, and $40–$125+ for STEM and professional tracks (checked 20 Oct 2026). Hours aren't guaranteed.

How long does DataAnnotation take to respond?

The FAQ says applicants are usually notified "within a few days" of finishing the Starter Assessment. If you never hear back, it says your application is still under review or wasn't accepted, and it can't reply to every applicant.

Can I retake the DataAnnotation Starter Assessment?

No. DataAnnotation says you can take it only once, with "no retakes or second chances."

Do you need a degree for DataAnnotation?

For general work it asks for a bachelor's degree or equivalent real-world experience. STEM tracks want a master's or PhD (or a bachelor's plus 10 years' experience), and professional tracks need licensed credentials.

How does DataAnnotation pay?

Through PayPal. You request a payout and the FAQ says deposits arrive within a few days. Your PayPal email must match the one you registered with.

Labels on this page follow our earnings labels. Nothing here is legal, tax or financial advice. Rules and prices change; check the linked sources before relying on them.

Related research notes

More answers

The weekly research note

One new note a week: costs, hours and earnings ranges, with sources. Subscribe and get our free guide, AI side income that actually pays (October 2026). Unsubscribe anytime. Privacy