AI Security is the fastest-moving corner of the security job market. The roles are new, the titles are inconsistent, and the candidate pool is small — which makes compensation hard to read. This guide pulls together published 2026 ranges so hiring managers can set realistic bands, and candidates can sanity-check their offers.
Quick answer: in the US, most full-time AI Security roles pay between $120,000 and $230,000 in base or total compensation for mid-to-senior individual contributors. Staff and principal roles, and anything at a frontier AI lab, go well beyond that — mostly through equity.
2026 salary ranges by role
| Role | Indicative US range | Notes |
|---|---|---|
| AI Red Teamer — junior | $80k – $120k total comp | Often moves over from pentesting or ML research |
| AI Red Teamer — mid-level | $120k – $170k total comp | |
| AI Red Teamer — senior | $170k – $220k total comp | |
| AI Red Teamer — staff / principal | $220k – $350k+ total comp | Frontier labs pay above this band |
| LLM / GenAI Security Engineer | $150k – $230k | Builder role: guardrails, agent permissions, secure RAG |
| Adversarial ML Engineer | $140k – $220k | Strong ML research background expected |
| AI Red Team Researcher | $180k – $280k+ | Research-heavy; publications matter |
For context, the U.S. Bureau of Labor Statistics reports a median of $129,180 for information security analysts overall (May 2025). AI Security roles sit well above that baseline because they demand two scarce skill sets at once: offensive or defensive security, and hands-on ML or LLM engineering.
Sources: infosec.qa — AI Red Teamer Salary 2026, Practical DevSecOps — Emerging AI Security Roles 2026, infosec.qa — AI Security Engineer Salary 2026, U.S. BLS. Ranges are indicative, compiled from public sources as of October 2026, and are not Fortium placement averages.
Five factors that move an AI Security offer
1. Employer type
Frontier AI labs and well-funded AI scale-ups compete hardest for this talent and lean heavily on equity. Large enterprises building their first AI security function usually pay strong base salaries but less upside. Security vendors adding AI features land in between.
2. Builder vs. breaker
Red teamers and LLM security engineers are often benchmarked together, but the markets differ. Engineers who ship production guardrails and secure agent architectures are in short supply because they must also be strong software engineers.
3. Depth of ML experience
A pentester who has learned prompt injection is valuable. A candidate who also understands training, fine-tuning, and evaluation design — and can attack the model, not just the application around it — commands a meaningful premium.
4. Public track record
In a field without established certifications, published research, conference talks, CTF results, and AI bug-bounty findings do the work that credentials do elsewhere. Candidates with visible work have more leverage.
5. Location and remote policy
Bay Area, New York, and Seattle bands sit at the top. Remote-first employers increasingly pay a single national band, which narrows the gap — but rarely closes it for the most senior roles.
For hiring managers: setting the band
- Benchmark the skill mix, not the title. “AI Security Engineer” can mean anything from a prompt-testing analyst to a principal engineer designing agent permission systems. Write down which of the two scarce skill sets you need most.
- Decide where you will flex. If base is capped, be ready to compete on equity, scope, research time, or conference budget — all of which this audience values.
- Move fast. Strong AI Security candidates rarely stay on the market for long. A slow interview loop costs more than an extra $10k of base.
For candidates: reading your offer
- Ask for the full breakdown: base, bonus target, equity (and its valuation basis), and refresh policy.
- Compare offers on total compensation over four years, not on base alone.
- If your role combines red-team and engineering responsibilities, benchmark against the higher of the two.
Get a benchmark for your exact role
Published ranges are a starting point. For a specific role, level, and location, request a free salary benchmark — we’ll send a tailored snapshot, typically within 2 business days. For role-by-role detail, see our hiring guides for the AI Red Teamer and the LLM Security Engineer.