The Layoffs Blamed on AI Are Moving in the Wrong Direction to Actually Be About AI

US tech employers announced 139,156 job cuts in the first half of 2026, up 83% from 76,214 over the same period last year. AI was explicitly cited in 101,743 layoff announcements this year, already dwarfing the 54,836 cited for all of 2025, according to Challenger, Gray & Christmas data. On its face, that looks like exactly the story being told: AI is automating jobs away.

The capex numbers tell a different story. Amazon cut 14,000 corporate roles in October 2025 and another 16,000 in January, while AWS grew 24%, its fastest quarter in thirteen years, and the company's planned capex rose from $131.8 billion to $200 billion. Meta cut roughly 8,000 staff, about 10% of its workforce, while shifting 7,000 people into AI roles and raising its 2026 capex guidance to $115 to $145 billion off revenue that grew 33% year over year. Combined, Amazon, Microsoft, Alphabet, and Meta are guiding toward roughly $700 billion in 2026 capex, nearly double what they actually spent in 2025.

The tell is in what's growing versus what's being cut

If AI were genuinely substituting for the labor being cut, the pattern to expect is layoffs concentrated in roles where automation is cheapest, alongside flat or declining infrastructure spending, since the whole point of automation is spending less. Instead, the same companies doing the cutting are nearly doubling AI capex, and the divisions being spared, like AWS, are the fastest-growing ones. Layoffs and AI investment are moving in the same direction, not opposite ones.

Sam Altman himself has conceded that companies blame AI for layoffs "whether or not it really is about AI." Marc Andreessen has offered a more mundane explanation: pandemic-era, near-zero interest rates drove years of overhiring, with some companies 25 to 75% overstaffed relative to actual need, and rate normalization is now forcing a correction that has nothing to do with automation.

Read together, the layoff and capex data support Andreessen's version more than the automation story. "AI" is functioning here as a shareholder-facing narrative that lets already-profitable companies convert an overdue overhiring correction into a growth story, framing headcount reduction as strategic pivot rather than correction. That doesn't mean AI has zero effect on any job anywhere. It means the specific companies and cuts generating the biggest headlines this year don't fit the automation explanation nearly as well as the spending data suggests they should.

Sources: TechCrunch · Forbes · CNBC · CFO Dive · Moneywise · SF Standard · Invezz