Spotify's Senior Engineers Haven't Written a Single Line of Code Since December — And the CEO Is Bragging About It
Spotify engineers stop writing code — and their CEO is calling it a victory. AI has fully taken over development at one of the world’s most-used apps, and Gustav Söderström said it out loud on a February 2026 earnings call without a hint of apology.
He didn’t frame it as a warning. He framed it as proof that Spotify’s AI-first bet is paying off — and the numbers back him up.
The trend of Spotify engineers stop writing code AI handles has accelerated faster than most industry analysts predicted heading into 2026.
Why Did Spotify Engineers Stop Writing Code? AI Explained
During a February 2026 earnings call, Gustav Söderström told investors that some of Spotify’s top engineers have effectively stepped away from manual coding entirely. Their new role is to guide, prompt, and oversee AI output — rather than type a single line themselves. Businesschief
This wasn’t buried in a footnote. Söderström quoted his own engineers directly: “When I speak to my most senior engineers — the best developers we have — they actually say that they haven’t written a single line of code since December.”
And he framed it as proof that AI-powered development at Spotify isn’t a side experiment. It’s mainstream. It’s working. And it’s only getting started.
Meet "Honk" — The Internal AI System Making This Possible
So how is a company with 751 million monthly active users shipping software without humans writing the code?
The answer is an internal system Spotify calls “Honk” — a platform that enables remote, real-time code deployment using generative AI, specifically powered by Claude Code from Anthropic. Mitsloanme
Here’s what that looks like in practice. Söderström described a real scenario on the call: an engineer on their morning commute opens Slack on their phone, tells Claude to fix a bug or add a new feature to the iOS app, and by the time they arrive at the office, a new version of the app has already been pushed to their phone for review and merge to production. Mitsloanme
That’s not a demo. That’s how Spotify is building its product in 2026.
Honk isn’t just a glorified autocomplete tool either. Spotify built “Fleet Management” starting in 2022 — a framework capable of performing code changes across hundreds or thousands of repositories simultaneously. The integration of the Claude Agent SDK only took place in July 2025. heise online What you’re seeing now is the result of years of infrastructure groundwork finally hitting its stride.
The Numbers Behind the Claim
Skeptics might wonder: is this just PR spin from a CEO trying to impress investors? The output data suggests otherwise.
In Q4 2025, Spotify reported that Premium subscribers increased 10% year-over-year to 290 million, while Monthly Active Users rose 11% to 751 million. Total revenue grew 13% year-over-year on a constant currency basis to €4.5 billion (approximately $5.3 billion). Gross margin improved to 33.1%, and operating income reached €701 million. Businesschief
Beyond the financials, Spotify shipped more than 50 new features and changes throughout 2025, including AI-powered Prompted Playlists, Page Match for audiobooks, and About This Song — all launched within the past few weeks of the announcement. TechCrunch
That’s a meaningful pace of shipping for a platform of this scale. And if AI is generating the code, the cost-per-feature curve is going in a very favorable direction for Spotify.
What "Not Writing Code" Actually Means for Engineers
Here’s where the conversation gets more nuanced — and more important.
When Söderström says engineers aren’t writing code, he doesn’t mean they’re watching Netflix at their desks. Spotify’s engineers have shifted from writing code to something closer to high-level product management — defining requirements, reviewing AI-generated code, and orchestrating system architecture while the AI handles implementation details. Techbuzz
The role hasn’t disappeared. It’s been abstracted upward.
Think of it like the transition from assembly language to high-level programming languages in the 1970s and 80s. Developers didn’t stop working — they started working at a higher level of abstraction. The machine still runs on the same fundamentals; humans just stopped managing every layer directly.
That’s what’s happening at Spotify. Engineers are becoming conductors rather than musicians. They set the tempo, choose the piece, and correct the performance — but the instruments play themselves.
Is This the Future of Software Development — Or a Ticking Time Bomb?
The reaction online has been sharply divided, and both camps have legitimate points.
The Optimist Case
Those excited about Spotify’s shift point to the productivity upside. According to a Deloitte study from 2025, only 11% of organizations use Agentic AI in production. 30% are exploring the topic, and 38% are conducting pilot projects. heise online Spotify is firmly in that rare 11% — and their output velocity suggests it’s paying dividends.
Söderström himself framed it unambiguously: “We foresee this not being the end of the line in terms of AI development, just the beginning.” TechSpot
For businesses and investors, AI-generated code means faster iteration, lower headcount costs over time, and the ability to ship features that would have previously required months of development cycles.
The Skeptic Case
The critics aren’t wrong either. Some commentators have argued this makes Spotify the most poorly managed software shop — pushing engineers to merge AI-generated code to production from their phones during a morning commute raises serious questions about oversight, testing, and quality control. Slashdot
There’s also the copyright elephant in the room. Courts have already weighed in on AI-generated work — the U.S. District Court for the District of Columbia ruled in 2023 that machine-created content cannot be copyrighted, even if a human was directing the machine. What that means for IP ownership of AI-written software remains murky legal territory. Slashdot
And then there’s the job question. It remains unclear whether this won’t ultimately affect developer job security — particularly for junior staff who lack the experience to effectively direct and review AI output. heise online Spotify already cut roughly 20% of its workforce between 2023 and 2025 across four rounds of layoffs. The timing is not coincidental.
What Sets Spotify's Approach Apart From GitHub Copilot and Other AI Coding Tools
Most companies using AI coding assistants are doing something far more incremental. GitHub Copilot suggests the next line. ChatGPT helps debug a function. These are productivity boosters grafted onto existing workflows.
Spotify is doing something structurally different. Rather than going all-in on external tools alone, Spotify built Honk — an internal AI system fine-tuned on Spotify’s specific codebase, architectural patterns, and development practices. Techbuzz
The result is an AI that doesn’t just understand generic code — it understands Spotify’s code. Its conventions, its infrastructure, its historical decisions. That’s a meaningful competitive moat that GitHub Copilot, trained on public repositories, simply cannot replicate out of the box.
Söderström made this point explicitly, describing a proprietary music dataset: “This is a dataset that we are building right now that no one else is really building. It does not exist at this scale. And we see it improving every time we retrain our models.” TechSpot
Spotify isn’t just using AI. It’s building AI that gets better the more Spotify uses it.
The moment Spotify engineers stop writing code AI systems take on is not a distant hypothetical — it's already the daily reality at one of the most-visited platforms on earth. What This Means for Software Engineers Right Now
If you’re a developer reading this, here’s the honest picture.
The junior developer pipeline is the most at-risk. Traditional entry-level roles — writing boilerplate code, fixing small bugs, adding minor features — are exactly the tasks AI handles first and best. If AI is handling those at Spotify, the on-ramp for new engineers entering the field gets steeper.
Senior engineers who can direct AI effectively become more valuable, not less. The engineers Söderström is bragging about aren’t unemployed — they’re doing more with less effort. But their value now rests on judgment, architectural thinking, and the ability to catch what the AI gets wrong. Those are skills that come from writing a lot of code first.
Quality assurance and security review become critical chokepoints. If AI writes all the code and humans review the output, the review process needs to be rigorous. The risk isn’t that AI writes bad code — it’s that humans stop knowing how to catch it.
The Competitive Pressure This Creates Across the Industry
Spotify’s announcement isn’t just a story about one company. It’s a signal that raises the bar across the entire streaming and tech landscape.
If one of the world’s most-used consumer apps can run its engineering on AI-generated code, platforms like Apple Music, Amazon Music, and YouTube Music face a stark competitive pressure to match that development velocity or fall behind. Techbuzz
Every product manager at a competing company is now asking their engineering leadership: “Why aren’t we doing what Spotify is doing?”
That question is going to move fast across the industry in 2026. And the companies that have been treating AI as a side project are about to feel the gap widen.
This is precisely why watching Spotify engineers stop writing code matters — it sets a new baseline expectation for what AI-assisted development looks like at scale.
Spotify's Bigger AI Ambition (That Nobody Is Talking About)
Buried beneath the coding headlines is a more strategic play that deserves attention.
Spotify’s co-CEO described the company as “the R&D department for the music industry” — but the R&D is going into AI coding tools, books, and video rather than new music monetization products. Complete Music Update That’s a significant strategic signal about where Spotify sees its future.
The company isn’t just using AI to build product faster. It’s using AI to reposition itself as a technology platform that happens to distribute audio — rather than a streaming service that happens to use technology.
Founder and Executive Chairman Daniel Ek has linked this to broader shifts: “The next wave of technology shifts — AI, new interfaces, wearables, new ways of interacting with content — these will reshape how people discover and experience audio and media.” Businesschief
Spotify is betting it can ride that wave by becoming an AI-first company before its competitors realize the race has already started.
Frequently Asked Questions
Q1: What does it mean that Spotify engineers haven’t written code since December? It means Spotify’s senior engineers have fully transitioned to prompting and reviewing AI-generated code rather than writing software manually. They describe requirements, AI systems like Claude Code generate the code, and engineers review and approve the output before deployment. When we say Spotify engineers stop writing code, it means Spotify’s senior developers have fully transitioned to prompting and reviewing AI-generated code rather than writing software manually.
Q2: What is Spotify’s “Honk” system? Honk is Spotify’s proprietary internal AI platform built on top of Anthropic’s Claude Code. It enables real-time, remote code deployment across Spotify’s infrastructure and is fine-tuned on Spotify’s specific codebase, making it more capable within Spotify’s environment than general-purpose AI coding tools.
Q3: Is Spotify’s AI-generated code safe to deploy to production? Spotify engineers still review and approve AI-generated code before it merges to production. However, critics have raised concerns about the rigor of review processes when engineers are approving code from their phones during commutes. The quality and security implications at scale remain an open question the company hasn’t addressed with public metrics.
Q4: Will AI replace software developers completely? Not imminently, but the role is changing fast. Junior developer tasks are most at risk as AI handles implementation work. Senior engineers who can effectively direct AI systems, review output for architectural and security issues, and make high-level system decisions remain highly valuable — possibly more so than before.
Q5: How does Spotify’s approach compare to GitHub Copilot? GitHub Copilot suggests code line-by-line within an IDE and acts as a productivity enhancement for human coders. Spotify’s approach through Honk and Claude Code is more agentic — engineers describe a goal, and AI completes entire features or bug fixes autonomously, then delivers results for review. It’s a fundamentally different level of delegation.
Q6: What are the legal implications of AI-written code? Ownership of AI-generated code is legally uncertain. A 2023 U.S. federal court ruling established that purely machine-generated content cannot be copyrighted. Whether code generated by AI with human oversight meets the threshold for copyright protection is an unresolved question that companies relying on AI-generated software will eventually need to navigate.
Q7: Which AI model is Spotify using for coding? Spotify is primarily using Claude Code, developed by Anthropic, integrated into their proprietary internal system called Honk. The Claude Agent SDK was integrated into Spotify’s Fleet Management framework in July 2025.
Conclusion
Watching Spotify engineers stop writing code AI systems generate instead is not a gimmick — it’s a live demonstration of what AI-first engineering looks like at 751 million users of scale. It’s a genuine, operational shift in how one of the world’s largest consumer tech platforms builds its product — and it’s working by every available metric.
The real question isn’t whether AI can write code. Spotify just proved it can. The real question is what it means for the people who built their careers on doing exactly that — and what the companies watching from the sidelines plan to do next.
This is the part of the AI revolution that feels abstract until a company with 751 million users makes it concrete. Spotify just made it concrete.
Want to stay ahead of how AI is reshaping software development and tech careers? [Insert internal link: AI’s impact on tech jobs] Subscribe to our newsletter for weekly analysis on the stories that actually matter.