Unlocking Elite Remote Careers: AI Safety Experts — English & Portuguese (Global) at Mercor ($29–$45/hr)
Company: Mercor · Pay: $29–$45/hr · Location: Remote — Global · Type: contractor · Languages: Portuguese · Posted: 2026-09-15
The artificial intelligence revolution is expanding faster than ever, and major AI labs are prioritizing robust safety testing over raw computing power. If you have native or near-native fluency in English and global Portuguese (excluding Brazilian Portuguese), an incredible fully remote career opportunity awaits you. Mercor is actively recruiting top linguistic talent for the AI Safety Experts — English & Portuguese (Global) position, offering an outstanding hourly pay rate ranging from $29 to $45.
Whether you want to break into the tech industry, leverage your multilingual capabilities, or enjoy location-independent work on your own schedule, this comprehensive guide covers everything you need to know about the role, its compensation, and how to successfully complete your application.
AI Safety Experts — English & Portuguese (Global)
Role Overview: Join an elite red-team data operation probing conversational AI models with adversarial inputs to uncover vulnerabilities and make next-gen AI safer.
Comprehensive Bilingual Article: English & Portuguese Side-by-Side
To help international professionals understand the nuances of this high-paying opportunity, the following deep dive is presented side-by-side in both English and Portuguese.
English Version
Introduction to AI Red Teaming
At Mercor, we believe the safest artificial intelligence model is one that has already been rigorously attacked and tested by human experts. We are building a specialized red team to probe advanced conversational AI models with adversarial inputs, surface systemic vulnerabilities, and generate critical safety datasets.
What You Will Do Daily
- Red Team Models: Test AI agents using jailbreaks, prompt injections, misuse cases, bias exploitation, and multi-turn manipulation techniques.
- Generate Human Data: Annotate model failures, classify vulnerabilities, and flag hidden risks.
- Apply Strict Taxonomies: Follow structured guidelines, benchmarks, and playbooks to maintain rigorous testing consistency.
- Document Findings: Produce clear attack case reports and datasets that clients can act on immediately.
Candidate Profile Requirements
- Prior red teaming experience (AI adversarial work, cybersecurity, or socio-technical probing).
- Curious and adversarial mindset: instinctively pushing systems to their absolute limits.
- Structured approach using frameworks rather than random hacks.
- Strong communication skills to explain complex risks to technical stakeholders.
Versão em Português
Introdução ao Red Teaming de IA
Na Mercor, acreditamos que o modelo de inteligência artificial mais seguro é aquele que já foi rigorosamente atacado e testado por especialistas humanos. Estamos construindo uma equipe especializada de red team para testar modelos avançados de IA conversacional com entradas adversariais, expor vulnerabilidades e gerar dados de segurança essenciais.
O Que Você Fará Diariamente
- Testar Modelos: Avaliar agentes de IA usando jailbreaks, injeções de prompt, exploração de viés e manipulação de múltiplos turnos.
- Gerar Dados Humanos: Anotar falhas de modelos, classificar vulnerabilidades e sinalizar riscos ocultos.
- Aplicar Taxonomias Estritas: Seguir diretrizes estruturadas, benchmarks e manuais para manter consistência nos testes.
- Documentar Descobertas: Produzir relatórios claros de casos de ataque e conjuntos de dados acionáveis.
Requisitos do Perfil do Candidato
- Experiência prévia em red teaming (trabalho adversarial de IA, cibersegurança ou sondagem sociotécnica).
- Mentalidade curiosa e adversarial: empurrar sistemas aos seus limites absolutos instintivamente.
- Abordagem estruturada usando frameworks em vez de hacks aleatórios.
- Excelentes habilidades de comunicação para explicar riscos complexos a partes interessadas técnicas.
💡 Note on Sensitivity: Projects may touch on topics such as bias, misinformation, or harmful behaviors. All work is text-based, and participation in sensitive projects is optional, supported by clear guidelines and wellness resources.
Contract Terms, Payment Structure, and Financial Growth
Mercor offers an independent contractor model designed for maximum flexibility and transparency. Key financial benefits include:
- Competitive Hourly Wage: Earn between $29 and $45 per hour based on your expertise and project allocation.
- Weekly Payouts: Reliable weekly disbursements processed securely through Stripe or Wise.
- Full Flexibility: Complete your tasks remotely on your own schedule with no rigid office hours.
- Lucrative Referral Bonuses: Earn up to $180 for every successful professional you refer to the platform, with no limits on total referrals.
Step-by-Step Guide to Completing Your Application Successfully
AI safety positions at leading technology firms are in massive demand. Follow these actionable steps to ensure your application stands out immediately:
- Highlight Adversarial or Analytical Experience: Make sure your resume highlights any background in cybersecurity, content moderation, policy evaluation, technical writing, or adversarial testing.
- Verify Linguistic Credentials: Clearly state your fluency in English and global Portuguese (noting that a separate track handles Brazilian Portuguese).
- Complete Screening Promptly: Mercor uses a streamlined interview and assessment pipeline. Complete all requested steps immediately to secure priority review as an early applicant.
Do not wait—cohort slots fill up rapidly. Take the first step toward a rewarding, high-paying remote AI career today!