25. Compensation Evidence
The earlier version of this page published broad entry, mid, and senior ranges assembled from secondary salary sites and adjacent-role estimates. That method could not support title-level comparisons, especially for new roles with inconsistent scope. This revision removes those estimates.
Current employer disclosures can establish a range for one role, level, location, and date. They are not market benchmarks:
| Posting observed in August 2026 | Published base range | Evidence boundary |
|---|---|---|
| OpenAI Applied AI Engineer, Enterprise, US | $197K to $278K, equity separate | Customer-facing enterprise scope at one employer |
| OpenAI Forward-Deployed Engineer, San Francisco | $162K to $280K, equity separate | One employer, one location, broad level range |
| OpenAI Backend Software Engineer (Evals), San Francisco and Seattle | $266K to $445K, equity separate | Product-domain eval infrastructure, not every evaluation role |
| Cisco AI Security Engineer, US locations | $150.3K to $195.2K | One posting with employer-specific research scope |
| Marvell AI Security Engineer, Santa Clara or Austin | $131.54K to $197K | One enterprise cyber-engineering role |
| BlackRock Senior AI Platform Engineer, Budapest | HUF 13.4M to 20.1M | Hungary, VP or Director level, employer-specific band |
For negotiation, compare live postings with the same country, city or remote policy, level, employer type, and compensation components. Do not convert one US lab range into a global role average.