Snap Fired 1,000 People. AI Now Writes 65% of Its Code. The Stock Jumped 7%.

Snap laid off approximately 1,000 full-time employees in April 2026, cutting 16% of its global workforce and closing over 300 open roles. CEO Evan Spiegel framed the restructuring as a direct consequence of artificial intelligence capability: the company’s engineering teams are now generating 65% of all new code with AI tools, and that shift means smaller human teams can accomplish what larger ones previously required. Snap’s stock jumped roughly 7% on the announcement.

Key Highlights

• Snap cut approximately 1,000 full-time employees and closed 300 open roles, reducing headcount by 16%
• AI tools now generate 65% of all new code written at Snap
• The company expects to save over $500 million annually by the second half of 2026
• US-based employees receive four months of severance, healthcare coverage, and equity vesting
• Snap stock rose approximately 7% following the announcement
• CEO Evan Spiegel described the moment as a “crucible” for the company and cited AI as enabling a fundamentally new way of working

The AI Workforce Shift Is No Longer Hypothetical

For the past two years, technology executives have warned that AI would transform software development teams. Snap’s announcement is one of the clearest examples yet of that transformation moving from forecast to fact.

When 65% of new code is generated by AI, the math on engineering headcount changes entirely. A team of 10 engineers with AI tools can produce the output that previously required 25 or 30. You do not need to eliminate the entire team. You need to right-size it to the new productivity baseline, and that right-sizing is what the 1,000 layoffs represent.

Spiegel’s internal communication described what he called a “new way of working,” in which AI agents handle large portions of routine development work while human engineers focus on architecture, product judgment, and quality review. The company is not abandoning software development. It is restructuring who does it and how.

The $500 Million Math

Snap expects the workforce reduction to cut annual expenses by more than $500 million, with savings materializing primarily in the second half of 2026. Engineering compensation, benefits, and associated costs account for the bulk of the savings.

The investor community responded immediately. A $500 million reduction in annual costs on a company of Snap’s size is a significant margin improvement. The 7% stock jump reflects markets pricing in a structurally leaner cost base, not just a one-time restructuring. If AI tools can sustain or grow output while the headcount stays lower, the efficiency gain is permanent rather than cyclical.

This is the dynamic that Jeff Bezos described as the “gigantic societal upside” of AI. The productivity gains are real. The disruption to individuals is also real. Both things are true simultaneously.

What Snap’s Move Signals for the Broader Tech Industry

Snap is not alone. The tech industry entered 2026 with a wave of AI-driven workforce restructurings across companies ranging from large platform businesses to mid-size SaaS providers. What makes Snap’s announcement notable is the specificity of the metric: 65% AI-generated code. That number gives other technology executives a public benchmark to reference.

If Snap can operate with 16% fewer engineers because AI handles 65% of code generation, the implicit question every technology board is now asking is: what is our equivalent number, and what does the right-sized team look like at that productivity level?

Sam Altman’s prediction that AI would reach and then exceed human-level capability at many cognitive tasks is playing out not as a distant speculative scenario but as a present-tense restructuring of how technology companies allocate labor. The velocity of the shift is faster than most workforce planning models accounted for.

The Human Cost

The 1,000 employees losing their jobs at Snap are real people. US-based workers will receive four months of severance pay, continued healthcare coverage through the severance period, accelerated equity vesting, and transition support including career placement services. International employees will receive packages in line with local legal requirements.

Four months of severance is above the minimum for most US tech layoffs, but it is not unlimited. In a job market where AI-related restructurings are happening across multiple companies simultaneously, the absorption capacity for displaced software engineers is a genuine open question. Companies like SAP have made similar moves, and the compounding effect of parallel layoffs across the industry is creating pressure on the senior engineering talent market that was not present a year ago.

Snap’s Broader Context

Snap has struggled with profitability since going public. The company has faced sustained pressure from TikTok and Instagram on its core social media business and has repeatedly restructured to reduce costs. The AI-driven restructuring in April 2026 is the most structurally significant of those efforts because it is not just cutting costs. It is changing the underlying production model for the company’s technology.

If the AI code generation tools maintain quality and output, Snap’s engineering efficiency will have structurally improved rather than simply shrunk. That is a meaningful difference from a traditional layoff cycle, where cost savings come at the expense of capability. The bet Spiegel is making is that AI tools are now good enough to preserve capability while reducing headcount.

The 7% stock reaction suggests investors believe him. The proof will come in product quality and engineering output over the next two to four quarters.

The TCB View

Snap’s 65% AI code generation number is the most important data point in this story. It is not a projection or an aspiration. It is what is happening right now at a major technology company. Enterprise AI adoption is moving from pilot programs to restructuring decisions. The transition from “AI will eventually change work” to “AI is changing work budgets and headcount today” has happened faster than most companies planned for. Snap is an early public data point in what will become a much larger pattern across the industry in 2026 and 2027.