Is Handshake AI Legit? What to Check Before You Apply

Handshake started as a campus job board and now also runs Handshake AI, a program that places graduates, researchers and professionals on AI-training and evaluation projects. People searching for it often ask whether it is legitimate. This guide lists what can be verified, what workers commonly report, and what to check yourself before you spend time on an application.

What Handshake AI is

Handshake describes Handshake AI as a flexible, project-based fellowship. In its own help center, it says the work is not a full-time employment opportunity. The work is contract-based and carries no guaranteed hours.

According to Handshake's help pages, typical tasks include writing or evaluating model responses, reviewing outputs, challenging models, labeling data and giving feedback. Work follows detailed project guidelines, and people set their own pace within each project. Handshake's FAQ for university partners also describes creating prompts in the fellow's field of expertise and comparing AI-generated responses.

The help center says you do not need to be a student. It says the program includes undergraduates, recent graduates, researchers, PhDs and working professionals. Projects are scoped to particular domains or backgrounds. Some third-party guides say the program is currently focused on people based in the United States, so check the current eligibility terms for your location.

What you can verify yourself

• The official help center. Handshake publishes articles on fellowship expectations and on payments, including pages about eligible earnings and payments processing. Read them before you apply.
• The project terms. Third-party guides say that terms such as hourly or per-task pay, time caps and whether onboarding is paid are set out per project. Read them before accepting a project.
• Independent reporting. Business and technology outlets have covered Handshake's data-labeling business and some worker disputes. These reports are a useful second view alongside the company's own pages.
• Review sites, with care. Glassdoor, Trustpilot and Reddit show worker experiences. Sample sizes are often small, and several guide sites that rank AI-training platforms earn affiliate income or sell courses.

What the process usually looks like

Handshake's official pages describe project-specific screening or assessments for many projects. Third-party guides describe a similar path: you sign up through the fellowship portal, provide your education and field, and may be asked for identity verification, a background check and an assessment. The assessment type varies by role and can involve critical thinking, prompt writing or subject knowledge such as math or coding.

Passing a screening does not mean you start earning. Guides and the help center describe matching: Handshake places people on projects based on qualifications and what each project needs. Future offers depend on past performance, fit and project needs. Reports of the wait for a first project range from days to months.

On payment, the help center points to separate payment articles. Pay varies by project and is shown by the platform, so rely on what the platform shows for each project rather than figures quoted on blogs. Third-party guides mention payout methods such as PayPal, Wise or bank transfer, but check which options apply to you.

Common complaints and limitations

Worker feedback is mixed, and the points below are reports, not findings.

• Inconsistent availability. Handshake itself says work fluctuates and can change within days, and that projects may end with limited notice.
• Waiting after approval. Some workers report long gaps between being accepted and receiving a project.
• Unclear unpaid time. Some guides say it is not clearly published whether onboarding or assessment time is paid on a given project. A legal complaint on this topic has been reported, and it is an allegation rather than a ruling.
• Account removals and pay disputes. News coverage from early 2026 reports that some contractors on certain projects said their accounts were suspended and pay withheld. Handshake reportedly gave reasons such as credential mismatches or task-time benchmarks. The articles describe workers' accounts and do not include a full company statement, so the facts remain contested.
• Quality scoring. Some reviews describe quality scoring as subjective and say a run of low scores can lead to removal from a project.
• Support. Some reviews say communication and support responses were slow.

None of this shows wrongdoing. It shows that the work is project-based and uncertain, and that disputes have been reported.

Red flags to watch for on any platform

• Any request to pay a fee, buy equipment or purchase training before you can work.
• Contact from someone who is not on an official domain, or messages asking you to move to a chat app.
• Requests for bank or identity details before you have confirmed you are on the official site.
• Promises of guaranteed hours, guaranteed income or fast approval.
• Pay terms that are missing or vague, or that change after you start.
• No clear way to ask about disputed or unpaid work.

Realistic expectations

Treat this as supplemental, flexible work rather than a stable job. Official pages say there are no guaranteed hours and that offers depend on fit, performance and project needs. Pay varies by project, and the platform shows the rate for each one. Competition for places is reported, and people with specialist credentials are generally the ones described as best matched.

Keep your own records of hours worked, including onboarding, and save the terms for each project. Set aside money for taxes if you work as a contractor.

Quick checklist before you apply

• Read the official help center pages on expectations and payments.
• Confirm you are on the official Handshake website and not a lookalike.
• Check that the eligibility terms cover your country and background.
• Find out what the project terms say about pay type, time limits and onboarding.
• Confirm you will not be asked to pay anything.
• Read a mix of recent worker reviews and news coverage.
• Keep a record of hours, messages and payment dates.
• Decide how you will cope if work pauses or ends.

Limitations of this guide

This guide is based on public pages and press and review coverage found at the time of writing. Handshake's terms, eligibility and project availability change, and some sources are third-party sites with their own interests. Reported disputes are allegations, and their outcomes may have changed. Always check the current official pages before you decide.

Last checked: October 2026

SOJI is an independent job board and is not affiliated with Handshake AI. SOJI does not list Handshake AI jobs.

Frequently asked questions

Is Handshake AI a real company?

Handshake is an established company known for its campus job platform, and Handshake AI is its program for AI-training and evaluation work. Its help center describes the fellowship and its expectations. Being a real company does not guarantee steady work, so read the project terms and recent worker reports as well.

Does Handshake AI guarantee work or hours?

No. Handshake's own help center says the fellowship is not full-time employment, work is contract-based, and there are no guaranteed hours. Availability can change within days, and offers depend on fit, performance and project needs. Treat it as flexible, supplemental income and plan for pauses between projects.

How much does Handshake AI pay?

Pay varies by project and is shown by the platform. Blogs quote many different ranges, and they often come from affiliate or self-reported sources, so they are not a reliable guide. Check the rate and payment terms stated for each project before you accept, and keep a record of your hours.

Do I need to be a student to apply?

According to Handshake's help center, no. It says the program includes undergraduates, recent graduates, researchers, PhDs and working professionals. Many projects are scoped to specific fields and use project-specific screening, so your background and subject knowledge matter. Some guides say eligibility is currently focused on the United States, so check the current terms.