Adversarial AI Testing Specialist (LLM Red Teaming) | $28.74/hr Remote

Crossing Hurdles; United States; Remote; USD 39,592; Full time

Job summary

A progressive technology company is seeking a remote AI Red Teamer to assess conversational AI models and identify vulnerabilities through red teaming exercises. Candidates should be fluent in both English and Brazilian Portuguese, with a background in cybersecurity or adversarial testing. This role involves producing actionable reports and datasets, and requires strong communication skills to interface with both technical and non-technical stakeholders. Competitive hourly compensation is offered.

Qualifications

• Native-level fluency in English and Brazilian Portuguese is required.
• Strong background in adversarial testing and cybersecurity.
• Excellent written communication skills for various stakeholders.

Responsibilities

• Conduct red team exercises on AI models and agents.
• Annotate failures and classify vulnerabilities effectively.
• Ensure consistent testing using structured benchmarks.

Skills

Fluency in English

Fluency in Brazilian Portuguese

Adversarial mindset

Clear written communication

Job description

Type: Hourly contract (Full-time or Part-time)

Compensation: $28.74/hour

Location: Remote

Role Responsibilities

• Red team conversational AI models and agents (jailbreaks, prompt injections, misuse cases, bias exploitation, multi-turn manipulation).
• Generate high-quality human data by annotating failures, classifying vulnerabilities, and flagging systemic risks.
• Apply structured taxonomies, benchmarks, and playbooks to ensure consistent adversarial testing.
• Produce reproducible reports, datasets, and attack cases customers can act on.
• Identify vulnerabilities missed by automated evaluation systems.

Requirements

• Native-level fluency in English and Brazilian Portuguese (required).
• Prior experience in AI red teaming, adversarial testing, cybersecurity, or socio-technical risk analysis.
• Strong adversarial mindset with structured, methodical testing approaches.
• Clear written communication for technical and non-technical stakeholders.
• Comfortable reviewing sensitive AI-generated content (guidelines and wellness support provided).