The Bureau of Labor Statistics (BLS) released its April 2026 Employment Situation on Friday 8 May: +115,000 nonfarm payrolls, above the 55,000 consensus forecast but well below March's revised +185,000. Unemployment held at 4.3%. Job gains concentrated in health care (+37,000), transportation and warehousing (+30,000), and retail trade (+22,000); no technology sector gains appeared in any growth category. February 2026 payrolls were revised further down to negative 156,000 from the already-negative negative 133,000.
February's revision deserves separate attention. February's negative reading was first published as negative 133,000; it has now been revised to negative 156,000, a further 23,000-job deterioration. The direction of revision, deeper into negative territory, runs against the surface-level narrative that April's 115,000 gain represented recovery.
The BLS GenAI workplace paper, which the bureau skipped on 14 April with no rescheduling announcement, remained unpublished at the time of the April employment release. More than four weeks have now passed with no public explanation and no announced return date. The paper is the BLS's own research on how generative AI is affecting workplace tasks and occupational exposure; its continued absence means the agency responsible for measuring US employment has not published its own assessment of the technology most frequently cited in corporate restructuring announcements. Challenger, Gray & Christmas's parallel April 2026 report, published 1 May, extended the cumulative AI-attributed job cut series past 107,094, recording 21,490 AI-attributed cuts in April alone representing 26% of total announced cuts.
The measurement gap is now the operative fact for every AI workforce policy debate in Washington. Stanford Digital Economy Lab's analysis found AI suppresses approximately one million annual US hires relative to the 2023 pace, running primarily through positions not opened rather than workers explicitly terminated. The BLS payroll series does not distinguish between positions not opened and positions eliminated. With the GenAI paper absent, the government's own tool for that distinction has been removed from the debate at the moment it carries most weight.
