AI security engineer salary
What AI security roles pay, from employer-posted base ranges at labs, big tech and security vendors to self-reported total compensation, with each figure labelled by source and date, and what moves pay.
6 MIN READ · UPDATED 4 NOVEMBER 2026
PRACTICE THIS:AI Red Teaming & Evaluation ·Agent, Tool & MCP Security
How to read pay figures
Each figure here is attributed. Posted ranges are taken from the salary disclosures that pay-transparency laws require in job ads; they show base pay for a single title, and one ad frequently covers several levels, so the upper end of a broad band reflects a senior hire rather than a normal offer. Self-reported total compensation from sites such as Levels.fyi includes stock and bonus and is not comparable with a posted base range.
Many postings close within weeks. The figures below were read in November 2026, and the table on this page renders from our company data, so it stays in step with the company pages.
What postings showed in November 2026
Frontier labs posted the highest base ranges. Anthropic's Red Team Engineer, Safeguards posting listed $320,000 to $405,000 base in San Francisco, and an OpenAI Preparedness posting for agentic AI threat research listed $293,000 to $405,000.
Big tech posted base ranges spread wider because each spans levels and locations: Microsoft's AI Security Researcher II listed $102,100 to $202,200 across the US, Amazon's Security Engineer on AGI Security $136,000 to $212,800, and Apple's Senior Security Engineer, AI Security $171,600 to $302,200 in Seattle. Security vendors sat in a similar band: CrowdStrike's agentic systems threat researcher role listed $140,000 to $215,000 and Wiz's AI Security Researcher $160,000 to $220,000.
Total compensation
At large public companies, stock can add as much again as base at senior levels. Levels.fyi's self-reported data, read on 4 November 2026, showed a median total compensation of $335,000 for Amazon security engineers and $250,000 for Google security engineers. These figures cover security engineers generally, not AI security roles specifically, and they are self-reported.
What moves pay
Level moves pay most: the same company's range can double between a mid-level and a staff posting. Employer family comes next, with labs posting above big tech base bands. Location matters within the US, where many postings list separate bands for San Francisco and New York. Specialisation matters less than level, but red team and research roles at labs posted at the top of the market in 2026.
Outside the US
First-party pay outside the US is scarce. In London, Anthropic posted £255,000 to £325,000 for a Staff Software Security Engineer, and the UK AI Security Institute posted £65,000 to £145,000 across levels, base plus a technical allowance, with a large employer pension contribution. We found too little first-party data for India to publish a range; single self-reported figures are not a salary.
Vetted bands, by company
A row appears only when a figure has a source we can cite, and the last column says what that source is. So the table holds 14 of the 34 companies on this site. 14 rows are employer postings: the range printed on an actual req. The others come from public aggregators, which rely on self-reported figures and skew toward the people who chose to share. Red team and product security roles at the same company often pay differently, so every row names the precise title the range was posted for. Comparing a range with a different title is the most common way salary research misleads.
| COMPANY | TITLE THE BAND COVERS | RANGE | BASE / TOTAL | SOURCE |
|---|---|---|---|---|
| Anthropic | Red Team Engineer, Safeguards | $320K – $405K | base | Employer posting |
| OpenAI | Security Researcher, Agentic AI Threats | $293K – $405K | base | Employer posting |
| Atlassian | Principal Enterprise Security Engineer | $221K – $289K | base | Employer posting |
| Stripe | Security Engineer | $173K – $260K | base | Employer posting |
| Apple | Senior Security Engineer, AI Security | $172K – $302K | base | Employer posting |
| Cloudflare | AI Security Research & Red Team Engineer | $166K – $208K | base | Employer posting |
| Wiz | AI Security Researcher | $160K – $220K | base | Employer posting |
| Scale AI | Strategic Projects Lead, Red Team | $151K – $189K | base | Employer posting |
| Meta | Security Engineer, AI Security Threat Analysis and Reporting | $147K – $208K | base | Employer posting |
| CrowdStrike | Sr. AI/LLM Threat Researcher, Agentic Systems | $140K – $215K | base | Employer posting |
| Amazon | Security Engineer, Artificial General Intelligence | $136K – $213K | base | Employer posting |
| Palo Alto Networks | Sr Staff Engineer Software (Backend, Prisma AIRS Red Teaming) | $126K – $205K | base | Employer posting |
| Microsoft | AI Security Researcher II | $102K – $202K | base | Employer posting |
| xAI | Application Security Engineer | $100K – $258K | base | Employer posting |
Anthropic: Employer-posted San Francisco base range on Anthropic's job board, read 4 November 2026 (board updated 21 October 2026). A London Staff Software Security Engineer posting listed £255,000 to £325,000.
OpenAI: Employer-posted San Francisco base range for a role posted 19 May 2026 and since closed, read via a job-board mirror on 4 November 2026.
Atlassian: Zone A US base range on Atlassian's careers site, read 4 November 2026. Zone B listed $199,800 to $260,850 and Zone C $184,500 to $240,875.
Stripe: Employer-posted US-remote base range on Stripe's careers site, read 4 November 2026.
Apple: Employer-posted Seattle base range for a role posted 12 January 2026, read via a job-board mirror on 4 November 2026. Equity is additional.
Cloudflare: Employer-posted New York base range on Cloudflare's job board, read 4 November 2026. Pay may vary by location.
Wiz: Employer-posted base range for a Remote-USA role with an application deadline of 29 September 2026, read via job-board mirrors on 4 November 2026; Wiz's own page did not display it.
Scale AI: Employer-posted base range (San Francisco, New York, Seattle) for a role posted 5 June 2026 and since closed, read via a job-board mirror on 4 November 2026.
Meta: Employer-posted Menlo Park base range read via a job-board mirror on 4 November 2026; the posting date could not be confirmed.
CrowdStrike: Employer-posted US base range for a role posted 9 June 2026 and since closed, read via a job-board mirror on 4 November 2026.
Amazon: Employer-posted base range (lowest to highest US geographic market) for a role posted 15 December 2025 and since closed, read via a job-board mirror on 4 November 2026.
Palo Alto Networks: Employer-posted Santa Clara base range for a role posted 22 May 2026 and since closed, read via a job-board mirror on 4 November 2026. Equity is additional.
Microsoft: Employer-posted US base range (San Francisco and New York: $133,800 to $219,200) for a role posted 30 July 2026, read via a job-board mirror on 4 November 2026.
xAI: Employer-posted Palo Alto base range on xAI's job board, read 4 November 2026 (board updated 29 October 2026). On-site five days a week.
No sourced figure yet
These employers hire for AI security, but we found no posting or aggregator range we could cite. We list them without a figure rather than estimate one.
Practise this material in these question topics:
FAQ
In November 2026, employer-posted US base ranges ran from about $100,000 at the bottom of wide big tech bands to $405,000 at the top of frontier lab red team postings. Total compensation at large companies adds stock and bonus on top of base.