A sourcing note before the analysis: the primary white paper has not been reviewed at the source level for this brief. What follows is based on secondary reporting by the North Dallas Gazette on the study’s findings. Every figure and conclusion attributed to the paper should be understood as reported, not independently confirmed. The Filter has flagged this for primary source follow-up. If you locate the UMD or LinkUp source document directly, the brief can be upgraded. The primary paper may be accessible at UMD’s institutional page or LinkUp’s research portal.
With that context established: according to the paper as reported, researchers analyzed approximately 155 million U.S. job postings dating back to 2018 to assess whether AI adoption is suppressing the volume of positions employers are advertising. The reported conclusion is that it isn’t. Total job posting volume has not declined as a result of AI adoption, per the study’s findings as described by the secondary source.
Two findings in particular are worth noting. AI-specific job postings reportedly grew to approximately 1% of all U.S. postings in 2025, up from a much smaller baseline, suggesting the category is still expanding, not contracting. Entry-level postings reportedly grew to roughly 6% of total listings, which challenges the intuition that AI is disproportionately eliminating junior roles.
Stop there, though, because the methodology has limits worth naming.
Job posting volume is a measure of employer demand as expressed through open positions. It doesn’t tell you what happened to wages in those categories. It doesn’t track how the content of a job changed, a posting for a “marketing analyst” in 2024 may require skills and produce outputs substantially different from a 2018 posting with the same title. And aggregate volume can be stable while the distribution of postings shifts significantly across industries, geographies, and seniority levels.
In other words, a study finding no aggregate decline is consistent with material displacement in specific sectors or roles. The Oracle and Block announcements this week, both of which cite AI as a rationale for cuts, are company-level events in specific industries. The UMD/LinkUp study is measuring something different: whether the total number of job openings across the full economy has gone down. Both things can be true simultaneously. Individual companies can reduce headcount citing AI while the aggregate market continues to generate new demand in adjacent or emerging categories.
That’s the methodological caveat the study’s secondary coverage tends not to foreground. TJS has covered the developer employment category shift separately, a case where sectoral displacement is visible even when aggregate demand holds.
What to watch: whether the primary white paper surfaces with a methodology appendix, and whether follow-up studies track wage and occupational category shifts alongside posting volume.