AI's Dual Impact: Job Cuts Juxtaposed with Rising Operational Costs

Tech executives are shedding thousands of jobs under the banner of AI restructuring, even as the mounting price of compute and token usage tests internal budgets.

AI & Culture Desk
Updated 5d ago 2 min read
AI's Dual Impact: Job Cuts Juxtaposed with Rising Operational Costs

Across corporate suites, the rhetoric surrounding artificial intelligence has settled into an uneasy paradox. On one hand, leadership teams are using an AI-first future to justify trimming headcount, trimming management layers, and flattening organizational charts. On the other hand, the financial realities of running these models are beginning to bite. A Financial Times analysis reported by TechCrunch shows that U.S. tech employers have eliminated nearly 140,000 jobs since the start of this year, with giants like Amazon, Oracle, Meta, and Microsoft responsible for nearly 50,000 of those reductions as billions flow toward infrastructure.

Yet the capital markets appear unconvinced by the rationale. The Financial Times found that businesses pointing to AI during layoff announcements underperformed the Nasdaq by almost 10% in the 30 trading days after disclosure. Recent cuts underscore the trend: Monday.com announced a 20% workforce reduction of over 600 people to support an AI-driven operating model, while co-founder Eran Zinman maintained in an employee memo that the shift was not designed to replace staff with software. Similar structural moves occurred at Intuit, which cut roughly 3,000 roles, and Cisco, which eliminated nearly 4,000 positions despite beating revenue expectations.

Where the human ledger shrinks, software and compute ledgers are rapidly swelling. A report from Wired reveals that corporate executives are now wrestling with tokenomics, the escalating cost of processing inputs and outputs through generative models. Around 300 companies addressed token costs on earnings calls in April and May, up from 93 during the same months a year ago. Cisco chief executive Chuck Robbins acknowledged on an earnings call that internal chatbot adoption has made token usage pretty crazy, while Box chief Aaron Levine characterized token budgeting as a heated priority.

The operational bills arrive unevenly across the landscape. Amplitude chief executive Spenser Skates noted that select engineers run tabs reaching thousands of dollars monthly on tokens alone. At 8x8, where 1,800 workers use Anthropic's Claude, chief transformation officer Joel Neeb reported the firm initially saved $5 million by canceling redundant subscriptions. However, with staff adopting Anthropic's Claude Opus 4.8 model—which costs nearly 1.7 times more than a predecessor deployed in February—the enterprise has initiated conversations about potential usage caps or downgrading tasks to cheaper model tiers.

For workers on the ground, the expectations shift continuously. At companies like Coinbase, which laid off about 700 workers while experimenting with single-person teams combining product and engineering tasks, leadership has urged staff to ship work in days instead of weeks. At other firms, employees are monitored through usage dashboards or pressured into mandatory adoption. While startups such as OpenAI and Anthropic continue to recruit, the broader operating layer is left balancing rising compute scrutiny against the very tools management insists they cannot work without.

Sources

  1. 01 The running list: major tech layoffs in 2026 where employers cited AI — TechCrunch
  2. 02 ‘Pretty Crazy’ Token Usage Is Testing Bosses’ Bet on AI — Wired