Remote | AI Safety & Adversarial Testing Specialist — $65–$80/hour
24-MAG · New York, NY
$65–$80/hrfrom the description
Jul 19, 2026
New York, NY
Jul 20, 2026
What this job asks for AI summary
A part-time, remote contract role focused on adversarial testing of frontier AI models. The work involves designing prompts to expose safety failures, evaluating model behaviour across high-risk domains such as cybersecurity, biosecurity, and misinformation, and producing structured vulnerability reports. It suits experienced professionals from AI safety, red-teaming, trust and safety, cybersecurity, or related investigative and scientific fields.
Senior level · 5+ years · Remote · Bachelor's required · Contract
We read this from the posting text with AI. Skim the description below before ruling yourself out.
Why we read it this way (9)
This role sits at an unusual intersection of AI safety, cybersecurity, and domain-specific risk research (biosecurity, misinformation, fraud, political content). No single SOC code fits cleanly; 15-1299 (Computer Occupations, All Other) is used as the best available catch-all, with 15-1212 (Information Security Analysts) as a meaningful runner-up given the red-teaming and vulnerability-analysis emphasis.
The posting lists eligible locations spanning most of Europe plus the United States. The caller has declared this fully remote with a national (US) candidate pool, so no metro is set.
The degree requirement states a bachelor's degree 'is required' but immediately notes that 'equivalent specialist experience in adversarial testing, safety research, or high-risk investigations may also be considered.' This is a soft gate; however, because the JD uses the word 'required' for the degree, Bachelors is recorded as the minimum — the equivalency clause is noted here for the hiring manager's awareness.
Skills such as 'adversarial prompt design' and 'AI safety evaluation' are not canonical tool names but are retained because the role's core deliverables are defined by these practices rather than specific named software tools. No concrete named platforms, frameworks, or languages are required by the posting.
The 'Nice to Have' section explicitly labels jailbreak testing, prompt engineering, RLHF, AI alignment, safety benchmarks, threat modelling, and responsible disclosure as preferred — all are marked accordingly.
Compensation is quoted as $65–$80 per hour; payments are made weekly via Stripe or Wise. This is an independent contractor (1099/corp-to-corp) engagement, not a salaried position.
Ignored 2 non-technology phrase(s) as skills (responsibilities/concepts, not named tools): adversarial prompt design, vulnerability analysis.
Posting is for a contract engagement — the market benchmarks below price full-time roles, so read the comp comparison with that in mind.
Caller marked this a fully-remote role — scored against the national candidate pool.
Read the full posting
The employer publishes the full description on their own site — read it there ↗. Or sign in to read it here — it's free, and it also lets you track this application.