The US labor market added 178,000 jobs in March, in response to the Bureau of Labor Statistics.
Abstract
- March job progress stayed modest whereas tech hiring remained weak and entry-level roles saved shrinking.
- AI use rose in places of work, however many employees reported rework, frustration, and decrease belief.
- Executives noticed features from AI instruments, whereas employees confronted errors and further checks each day.
The information confirmed restricted change from the prior month, whilst corporations saved speaking about AI-led progress and higher office effectivity.
That hole has saved consideration on whether or not AI is lifting hiring and output as promised. Latest labor, office, and trade reviews present a extra combined image, particularly in tech and entry-level roles.
Most job progress in March got here from healthcare, development, transportation and warehousing, and social help. Healthcare added 76,000 jobs, whereas development gained 26,000 and transportation and warehousing added 21,000.
The BLS information didn’t present the identical energy in tech-linked areas. Computing infrastructure suppliers and net search portals confirmed little motion, whereas laptop techniques design and associated companies misplaced 13,000 jobs throughout the month.
That sample stands in distinction to public claims that tech hiring is recovering. Marc Andreessen mentioned fears about AI-led job losses have been overblown and shared information exhibiting extra job openings at tech companies.
However openings don’t at all times result in hiring. The March labor figures confirmed that the strongest hiring got here from sectors outdoors core tech, whereas associated digital companies stayed flat or moved decrease.
A latest Goldman Sachs report, cited by Fortune, mentioned AI minimize about 16,000 jobs monthly over the previous 12 months. On the identical time, a 2025 SignalFire examine mentioned new graduate hiring had dropped 50% from ranges seen earlier than the COVID-19 pandemic.
SignalFire mentioned, “The door to tech as soon as swung large open for brand spanking new grads. Immediately, it’s barely cracked.” The report linked that shift to smaller funding rounds, leaner groups, fewer graduate applications, and rising AI use.
Goldman Sachs additionally warned that employees pushed out by expertise usually transfer into extra routine jobs. The report mentioned this shift can cut back the worth of their present expertise and weaken labor outcomes for years.
That concern has widened the controversy round AI and employment. Whereas some leaders nonetheless anticipate long-term features, latest information has saved consideration on present hiring patterns and who bears the price of the change.
Employee expertise doesn’t match government optimism
Executives proceed to report robust assist for AI instruments. Harvard Enterprise Evaluate mentioned 80% of leaders use AI weekly, whereas 74% reported optimistic returns from early deployments.
Employees reported a special expertise. Mercer mentioned 43% of employees discovered their jobs extra irritating, whereas Workday mentioned practically 4 hours are misplaced fixing AI output for each 10 hours of claimed effectivity features.
Harvard Enterprise Evaluate additionally pointed to “workslop,” described as content material that appears polished however lacks substance. Researchers mentioned 41% of employees had seen this type of output, with every case including nearly two hours of rework.
Workday mentioned solely 14% of respondents “persistently obtain net-positive outcomes from AI use.” That end result suggests many workplaces are nonetheless coping with errors, further evaluate, and weak belief in outputs.
OpenAI warns coverage might lag behind change
The divide between government use and each day employees expertise might come from how groups use the instruments. Harvard Enterprise Evaluate mentioned senior leaders usually apply AI to technique, drafting, and synthesis, the place the techniques are likely to carry out higher.
For routine operations that want regular accuracy, outcomes seem much less dependable. Brian Solis of ServiceNow known as this burden an “AI tax,” which he described as “Extra checking. Extra rework. Extra anxiousness.”
OpenAI has additionally acknowledged that AI is altering employment. Its coverage concepts included broader healthcare protection, retirement financial savings assist, and a brand new industrial agenda.
The corporate mentioned its proposals are early and meant to start dialogue. It additionally warned, “Except coverage retains tempo with technological change, the establishments and security nets wanted to navigate this transition may fall behind.”
