Snap Fires 1,000 as AI Now Writes 65% of Its Code

Snap cut 16% of its workforce on April 15, citing AI-generated code as the direct cause. The stock jumped. The workers left. Here is what the company's own numbers actually say.

Snap Fires 1,000 as AI Now Writes 65% of Its Code

On Tuesday, Snap CEO Evan Spiegel announced he was cutting 1,000 employees - 16% of the company's full-time staff. He framed it as a transformation, not a retreat. The stated trigger: artificial intelligence now produces more than 65% of Snap's newly written code, and the company no longer needs as many people to write the rest.

TL;DR

  • Snap removed 1,000 jobs and closed 300 open roles on April 15
  • CEO Spiegel cited AI handling 65%+ of new code as the direct justification
  • Shares jumped 5-8% on the news; the stock is still down 31% year-to-date
  • Activist investor Irenic Capital had been pushing for cost cuts before the announcement
  • Snap is the latest in a string of tech companies citing AI to justify mass layoffs

What Snap Actually Said

Spiegel sent a memo to staff that was anything but quiet. He called the moment "a crucible" for the company - a word choice that signals something more than a routine restructuring.

The Crucible Moment

"Rapid advancements in artificial intelligence enable our teams to reduce repetitive work, increase velocity, and better support our community, partners, and advertisers."

That's the official line. But Spiegel had said something more revealing a few months earlier: AI is "probably the best thing that's ever happened" to Snap. He described competitive history as "like trench warfare with monopolies for 15 years" and predicted that "18 months from now, the way that Snap operates will be completely different."

The memo points to small teams "leveraging AI tools" that have already driven progress on Snapchat+, ad platform performance, and infrastructure. The company says it's shifting to "smaller, more focused teams powered by AI agents" - a phrase that's doing a lot of work in a very short sentence.

An open-plan technology office with rows of workstations and large monitors The kind of office Snap will need fewer of, if Spiegel's productivity claims hold. Source: unsplash.com

The 65% Coding Claim

The most specific number in the announcement is also the most consequential. Snap says AI creates more than 65% of its newly written code. That's not a projection or a target - Spiegel presented it as a current operational fact.

This places Snap with a small but growing cohort of companies reporting concrete automation of software development. Getting started with AI coding assistants used to be a beginner's exercise; these tools are now running production systems at scale. A figure above 50% for code generation - at a company with nearly one billion monthly active users - is worth pausing on. If accurate, it means Snap's engineering output has already been restructured around machines doing most of the writing.

The Financial Picture

The numbers tell a more complicated story than a clean AI-driven transformation.

MetricValue
Jobs cut1,000 (16% of workforce)
Open roles closed300+
Expected annual savings$500M+ by H2 2026
Q1 2026 revenue forecast~$1.53B (up 12% YoY)
2025 net loss$460 million
Stock move on layoff day+5.8%
Stock YTD performance-31%
Monthly active users (Q4 2025)946 million

Snap lost $460 million in 2025. The company's stock has shed nearly a third of its value this year. The $500 million in projected annual savings won't close that gap on its own, but investors clearly think it helps - shares climbed on the news even as a thousand workers were told to go home.

A developer writing code on a laptop in a modern workspace Snap says AI now does more than 65% of this. The question is whether that figure reflects reality or aspiration. Source: unsplash.com

What It Does Not Tell You

The Activist Pressure

Irenic Capital Management, which holds roughly a 2.5% economic interest in Snap, had been pushing the company to restructure before this announcement. The fund had specifically urged Snap to either divest or shut down its AR Specs unit, which has consumed over $3.5 billion in investment and loses approximately $500 million annually.

The timing deserves attention. A cost-cutting push from an activist investor, followed by a layoff announcement citing AI, is a pattern worth watching. The question of whether these cuts were planned before or because of AI is not one Spiegel answered directly.

A Second Round in 18 Months

This is Snap's second significant layoff round since October 2024, when the company cut roughly 10% of its workforce. Two rounds in less than two years suggests operational instability that AI framing alone doesn't fully explain. The first round wasn't sold as an AI transition. This one is.

The Broader 2026 Pattern

Snap isn't an outlier. More than 80 technology companies have cut over 71,000 jobs in 2026 so far, and AI is appearing with increasing frequency in the justifications. Atlassian laid off 1,600 workers in March and explicitly framed it as redirecting resources toward AI. Meta, Oracle, and Block have done similar. Anthropic's own research found that computer programmers face 75% AI exposure on a metric that tracks actual tool usage, not theoretical risk.

The 2025 Challenger data showed 55,000 U.S. workers had AI cited as the reason for their termination - more than twelve times the figure from two years prior. Economists at Oxford and Yale noted that companies may be using AI as a convenient framing for decisions made for other reasons. Snap's announcement doesn't settle that question.


When Spiegel said there is "no moat in software," he was warning that features get copied quickly in an AI-powered world. The inversion that follows from that logic is this: if software advantages are temporary, the workers who build software are the first to feel it. Snap's 65% coding figure, if it holds up, is the clearest number yet from a consumer-scale company on just how much of that writing has already moved to machines. The 1,000 workers who learned about it by memo on Tuesday probably already knew it was coming.

Sources:

Snap Fires 1,000 as AI Now Writes 65% of Its Code
About the author Senior AI Editor & Investigative Journalist

Elena is a technology journalist with over eight years of experience covering artificial intelligence, machine learning, and the startup ecosystem.