Coinbase cut roughly 700 employees. Meta slashed about 8,000. Cisco dropped 4,000. All in May 2026, all citing AI-driven efficiency as a primary factor.

But a Stanford research paper, led by economist Erik Brynjolfsson and released in August 2025, tells a more complicated story. The actual economic data says AI’s impact on overall employment remains minimal. The caveat is that “overall” is doing a lot of heavy lifting in that sentence.

What the data actually shows

The paper, titled “Canaries in the Coal Mine,” analyzed ADP payroll data to track real employment shifts since late 2022.

For most workers, the numbers barely moved. Experienced professionals in roles where AI augments rather than replaces their work saw stability or even growth in employment.

But for one specific demographic, the picture looks different. Workers aged 22 to 25 in occupations heavily exposed to AI, including entry-level software development, customer service, and accounting, experienced a 13% relative decline in employment. By some metrics, that decline reached as high as 16%.

The Stanford AI Index Report released in early 2026 reinforced these findings, painting a picture of narrow displacement concentrated in entry-level positions rather than economy-wide disruption.

The tech layoff wave in context

The broader tech industry shed over 90,000 jobs in 2026.

Coinbase’s reduction of approximately 700 jobs represented 14 to 15% of its workforce. The exchange explicitly pointed to increased operational efficiency driven by AI as a factor. Meta and Cisco told similar stories with their respective cuts of 8,000 and 4,000 positions.

The Stanford data suggests roles that involve fully automatable tasks are genuinely shrinking, while roles where AI functions as a productivity multiplier are holding steady or growing.

What this means for crypto investors

When one of the largest US-based exchanges cuts 14 to 15% of staff while citing AI efficiency, it signals something specific about how crypto companies are likely to operate going forward. Leaner teams powered by AI tools could meaningfully improve profit margins for exchanges and platforms, with lower fixed costs creating more resilience during downturns.

The broader pattern of AI-driven workforce optimization across tech also has second-order effects on the crypto ecosystem. A significant portion of crypto’s developer talent pipeline comes from exactly the demographic Stanford identified as most affected: young, early-career tech workers. If entry-level hiring continues to contract in AI-exposed fields, the pool of junior developers available for blockchain projects could shrink.