cover image
Handshake

Handshake

joinhandshake.com

1 Job

1,155 Employees

About the Company


Handshake is the career platform for Gen Z. With a community of over 17 million students, alumni, employers, and career educators, Handshake's network is where career advice and discovery turn into first, second, and third jobs.

Nearly 1 million companies use Handshake to build their future workforce--from Fortune 500 to federal agencies, school districts to startups, healthcare systems to small businesses.

Handshake is built for where you're going, not where you've been.

Listed Jobs

Company background Company brand
Company Name
Handshake
Job Title
Red Teaming Domain Expert - AI Training (Contract)
Job Description
Job title: Red Teaming Domain Expert – AI Training (Contract) Role Summary: Independent remote contractor who designs adversarial prompts to stress‑test and uncover vulnerabilities in large language models (LLMs). Works with a global team of engineers, researchers, and safety experts to evaluate model robustness and contribute to AI safety improvements. Expectations: - 10–20 hours per week, variable commitment. - Fully remote contract position (no visa sponsorship). - Requires flexible schedule, including possible evenings or weekends. Key Responsibilities: - Craft creative, adversarial prompts and scenarios aimed at breaking AI guardrails. - Identify and bypass safety filters, restrictions, and defenses. - Explore edge cases that trigger disallowed, harmful, or incorrect outputs. - Document experiments comprehensively—methods used, outcomes, and rationale. - Review and refine adversarial content produced by fellow team members. - Collaborate with engineers, tutors, and researchers to analyze findings and strengthen defenses. - Stay updated on new jailbreak techniques, attack methods, and evolving model behaviors. - Handle potentially disturbing content (violence, explicit material, hate speech) with professional composure. Required Skills: - Hands‑on experience using multiple LLMs. - Intuitive prompt crafting and familiarity with jailbreak or evasion strategies (preferred). - Creative, adversarial problem‑solving mindset. - Clear, thoughtful written communication. - Ability to tolerate emotionally heavy or graphic material. - Curiosity, persistence, and resilience amid frequent experimental failures. - Strong ethical judgment, separating adversarial tactics from personal values. - Self‑directed, collaborative, and comfortable in feedback‑dense environments. Required Education & Certifications: - No formal degree required. - Background in creative fields (writing, visual arts, game design, improv, niche internet culture) is a plus. - Prior red‑teaming, moderation, or adversarial testing experience is extra credit. - Familiarity with safety, trust & safety, or digital security concepts is advantageous.
United states
Remote
11-12-2025