Data current as of September 4, 2026. Numbers on this page snapshot the JobShift dataset on the date above; the AI layoffs view refreshes continuously as new WARN filings land.
The September 2026 cut
Uber announced a reduction affecting more than 3,000 employees in early September 2026, reported at approximately 10 percent of its global workforce and described by Forbes as the company's largest layoff wave since the pandemic. State WARN filings confirm 93 affected workers in Seattle and 1,396 in Chicago; the remainder of the headline figure has not yet appeared as individual state filings, which is normal — state WARN offices typically lag a national announcement by days to weeks.
Forbes reported the reduction as part of a push toward a "leaner organization," alongside a return-to-office policy limiting remote work to roughly 1 percent of staff, and framed cost pressure from prior AI spending and reinvestment in AI-driven initiatives such as robotaxis as contributing factors — without describing the layoff itself as AI-driven. JobShift's classifier reached the same conclusion independently: the September cut is not labeled AI-attributed, because the classifier requires causal language directly tying a layoff to AI in corroborating coverage, and neither Uber's own framing nor the press coverage on file meets that bar for this event.
Six waves since 2018
The JobShift dataset holds Uber layoff records going back to 2018. Excluding the September 2026 event, five prior waves appear on file, each counted from state WARN filings rather than a national headline figure:
| Filing date(s) | Location(s) | Workers (WARN-confirmed) | AI-attributed |
|---|---|---|---|
| 2018-05-29 | Tempe, AZ | 296 | No |
| 2019-09-10 | Palo Alto, CA | 62 | No |
| 2020-05-07 to 2020-05-18 | Phoenix AZ; Long Island City and Brooklyn, NY; Santa Clara County, CA | 513 | No |
| 2026-06-03 to 2026-06-12 | San Francisco, CA | 32 | No |
| 2026-07-22 | San Francisco, CA | 41 | Yes |
The 2020 wave lands during the initial COVID-19 disruption to ride-hailing demand and is the largest of the pre-2026 waves. The July 2026 wave is the only one on file tied to AI in corroborating coverage — reported at the time as a roughly 10-percent cut to customer-service roles citing AI adoption. At 41 confirmed workers, it is the second-smallest wave in the table, larger only than the 32-worker San Francisco filing from the previous month.
One AI-attributed wave in eight years
Across six layoff waves spanning 2018 through September 2026, exactly one carries an AI-attributed label in the JobShift dataset: July 2026, a customer-service reduction confirmed at 41 workers via California WARN filings. The September 2026 wave does not carry that label. Its state WARN filings confirmed so far — 93 in Washington and 1,396 in Illinois, 1,489 combined — already exceed the combined WARN-confirmed total of all five prior waves on file (944), before accounting for Uber's headline figure of more than 3,000 reported nationally.
This does not mean AI played no role in Uber's broader 2026 cost structure. It means the causal-attribution bar the classifier applies — a layoff has to be described as AI-driven, not merely adjacent to AI spending or strategy, to receive the label — was not met for the September event, matching how the reporting on it was itself framed. The classifier methodology documents this attribution logic in full.
A clustering limitation worth disclosing
JobShift groups layoff filings from the same company into clusters when they fall within a bounded window of each other, so that a single restructuring producing dozens of per-site WARN filings counts once rather than once per site. Uber's July and September 2026 waves fall inside that window relative to each other and are grouped into a single cluster internally, despite one being AI-attributed and the other not.
This is a real limitation of cluster-based counting, not specific to Uber: a company with two separate, differently caused layoff decisions close enough in time can be undercounted as a single event by any cluster-count metric, including the ones JobShift uses elsewhere on this site. The companies doing both AI hiring and AI layoffs piece, which lists Uber's AI-attributed cluster at approximately 41 workers as of its August 20, 2026 snapshot, predates the September wave and has not been revised to reflect it, consistent with that page's stated snapshot date.
What the pattern does and does not support
The narrative that AI is now a dominant, explicit driver of Uber's headcount reductions is not supported by the eight-year record on file: five of six waves predate or fall outside any AI attribution entirely, and the largest wave to date, arriving after AI attribution had already appeared once, was not itself labeled AI-driven by either Uber's framing or independent classification.
The narrative that AI is irrelevant to Uber's 2026 cost decisions is also not supported: AI spending and AI-driven reinvestment appear as stated contributing context for the September wave, even without meeting the bar for a causal AI-attributed label. Both the July and September 2026 waves happened at a company simultaneously investing in AI-role hiring, a pattern documented across a broader set of firms in the companies doing both analysis.
Live data
- The AI layoffs view shows current AI-attributed layoffs with company and state detail.
- The methodology page documents the AI-attribution logic applied to layoff events.