Skip to content
Browse jobs
Menu

job type

Remote Trust and Safety Jobs in AI

5 open jobsmost pay $53–110/hr1 platform hiringupdated 8 days ago

These are live remote trust and safety jobs from the platforms we track, checked every 24 hours.

Country not listed? Pick ‘Any country’ and check each job’s ‘Open to’ line.

Hiring now

Open jobs

eligibility

Do I qualify?

What you need

  • Comfortable reviewing difficult content
  • Careful, consistent judgement against a written policy
  • Policy, legal or security background for senior roles
  • Pass a policy-labelling test
Check where I can work

Countries these jobs accept

Open worldwide
4 of 5
Open in United States
5 of 5
Open in United Kingdom
4 of 5
Open in India
4 of 5
Open in Kenya
4 of 5
Open in Philippines
4 of 5

pay

What trust and safety jobs pay, by platform

Median and range in US dollars per hour, from the live jobs on this page.

  • Mercor

    $45–120/hr · 5 open

    $70/hr median

  • Pay for 5 live jobs with listed pay, by platform. Data as of 18 Oct 2026.

platforms

Platforms hiring for these jobs

Sorted by Legit Score, our 0–100 rating of how safe each platform is to work for.

  • MercorLegit · not rated yet

    Experts in law, medicine, finance and coding get matched to AI-training contracts after a résumé review, a 20-minute AI interview and assessments.

    Pay
    $15–500/hr
    Open roles
    297
    Countries
    Most countries

the work

What the work is

Remote trust and safety jobs in AI pay reviewers to check model output for harm, abuse or policy breaches. Red-teaming roles pay you to try to get a model to break its rules. Across the 5 live jobs on this page that list pay, most pay $53–110/hr, and the median is $70/hr. Expect some of the material to be disturbing, and read each listing's description closely before applying.

Words you will see on this page

  • Red teaming (trying to make an AI say or do something harmful, so the gaps can be fixed)
  • Legit Score (our 0–100 rating of how safe a platform is to work for, based on published inputs)

faq

Real questions, short answers

What is red teaming in AI?

Red teaming is deliberate stress-testing: you try to make an AI model produce harmful or rule-breaking output, then you document the method. Companies run these exercises to surface weaknesses before release.

Will I see disturbing content?

Sometimes. Good listings spell that out up front. Before accepting, ask what wellbeing support is offered and whether you can opt out of certain content.

How much do remote trust and safety jobs pay?

Across the 5 live jobs on this page that list pay, most pay $53–110/hr, and the median is $70/hr.

Do I need experience?

Generalist review roles often provide training. Senior policy and red-teaming positions expect experience in content moderation, security, law or a related field.