AI Safety Expert, English – Indonesian

This is a fully remote position, open to applicants in United States.

📋 Description

• Engage in red-teaming of conversational AI models and agents through the use of jailbreaks, prompt injections, misuse scenarios, bias exploitation, and multi-turn manipulation.

• Create human data by annotating failures, classifying vulnerabilities, and identifying systemic risks.

• Implement taxonomies, benchmarks, and playbooks to ensure consistency in testing.

• Generate reproducible reports, datasets, and attack cases for clients.

• Investigate AI systems to reveal vulnerabilities that automated tests may overlook.

• Broaden evaluation coverage and minimize unforeseen issues in production.

• Participate in projects aimed at training and improving frontier AI systems.

⛳️ Requirements

• Proficient in both English and Indonesian is essential.

• Previous experience in red-teaming related to AI adversarial activities, cybersecurity, or socio-technical probing is required.

• Capability to probe systems in an adversarial manner and challenge them to their limits.

• Proficiency in utilizing frameworks, taxonomies, benchmarks, and structured playbooks.

• Skill in clearly communicating risks to both technical and non-technical audiences.

• Flexibility to adapt across various projects and client needs.

• Must hold independent contractor status.

• Required to operate without access to confidential or proprietary information from any other employer, client, or organization.

• H1-B and STEM OPT candidates are not eligible.

🏝️ Benefits

• Work remotely from anywhere.

• Enjoy a flexible schedule tailored to your needs.

• Receive weekly payments through Stripe or Wise.

• Opportunities to engage in higher-sensitivity projects on an optional basis.

• Access to clear guidelines and wellness resources for projects involving sensitive content.

• Competitive compensation package.

• Chance to collaborate with top-tier researchers in the field.

• Earn up to $100 for every successful referral.