# The AI Jobs Report (2026)

Data through June 2026.

Across 204,223 US tech postings, 38.7% of
descriptions mention AI but only 22.3% list a concrete AI skill and
11% mention an LLM. The gap is "AI-washing": AI as language, not
a requirement. The real AI demand is narrower — and better paid.

## Headline numbers
- Descriptions mentioning AI: 38.7% (lower bound, 25% sample n=81,445)
- Postings listing an AI skill: 22.3% (n=203,310)
- Descriptions mentioning an LLM: 11%
- AI Engineer median midpoint: $160K vs Software Engineer $137K (+16.8%)
- Machine Learning Engineer median midpoint: $162K

## AI tools rising fastest (W11–16 → W17–22)
- Cursor: 0.1% → 0.3% (~3.2×)
- Pinecone: 0.1% → 0.2% (~3.1×)
- Claude Code: 0.1% → 0.4% (~3×)
- Firebase: 0.1% → 0.2% (~2.4×)
- Claude: 0.1% → 0.3% (~2.3×)

Honesty note: requiredSkills is AI-extracted; an extraction-prompt change mid-period
collapsed some soft-skill strings (~25×). Tool-name growth is more trustworthy than
soft-skill decline; treat exact multiples as directional.

Source: SoviaJobs Job Market Research. Salary = platform-estimated posted ranges.
Description figures are lower bounds (truncated text, 25% sample).
Free to use with citation + link to https://www.soviajobs.com/market-research.