Spotify just said the quiet part out loud.
During the company's Q4 2025 earnings call, CEO Gustav Soderstrom told investors that Spotify's most senior engineers -- the best developers in the building -- have not manually written a single line of code since December 2025. Instead, they generate code with AI tools and supervise the output.
"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," Soderstrom said. "They actually only generate code and supervise it."
This is not a research experiment or a limited pilot. It is a public statement from the CEO of one of the world's largest streaming platforms, a company with over 10,000 employees and 675 million monthly active users, telling Wall Street that his engineering team has fundamentally changed how it builds software.
What Spotify Is Actually Doing
Spotify has been investing heavily in AI-assisted development tools, including its internal system called Honk and external tools like Claude Code from Anthropic. Senior engineers at the company now spend their time defining what needs to be built, prompting AI to generate the code, and then reviewing the output rather than typing it themselves.
Soderstrom framed this as an existential imperative. "There is going to have to be a lot of change in these tech companies if you want to stay competitive, and we are absolutely hell-bent on leading that change," he said. "Engineering practices, product practices, and design practices will change."
He also acknowledged the pace is relentless: "The things you build now may be useless in a month."
The result, according to Spotify, is that the company will "produce massively more software, up until our limiting factor is actually the amount of change that consumers are comfortable with." In other words, the bottleneck is no longer engineering capacity. It is how fast users can absorb new features.
The AI Fatigue Problem Nobody Talks About
Soderstrom's optimism does not tell the full story. The same week, a viral essay by software engineer Siddhant Khare described a growing phenomenon he calls "AI fatigue" -- the exhaustion that comes from endlessly reviewing and fixing AI-generated code.
"Every time it feels like you are a judge at an assembly line and that assembly line is never-ending," Khare wrote. "You just keep stamping those PRs."
This is a critical nuance that businesses rushing to adopt AI coding tools need to understand. AI does not eliminate work. It changes the kind of work. Engineers go from writing code to reviewing code at a much higher volume. If your review processes are not built for that volume, you end up with burned-out senior engineers rubber-stamping pull requests -- the exact opposite of the quality improvement you were hoping for.
What This Means for Small and Mid-Size Businesses
If you run a business that builds or maintains software -- and in 2026, that includes most businesses -- the Spotify story is not just tech industry gossip. It is a signal about where development is heading at every scale.
1. The Role of "Developer" Is Changing Fast
Your developers are becoming supervisors of AI output rather than authors of code. This is the shift we explored in our post on agentic engineering displacing vibe coding. The trend has now moved from industry thought leaders to Fortune 500 earnings calls. It is real.
2. You Need Better Code Review, Not Just Better Code Generation
The lesson from Spotify's internal experience -- and from the AI fatigue backlash -- is that generating code is the easy part. Reviewing it at scale is where companies stumble. Before you deploy AI coding tools across your team, invest in:
- Automated testing pipelines that catch issues before humans even see the code
- Clear prompting standards so AI output is consistent and predictable
- Tiered review processes where junior reviews handle routine changes and senior engineers focus on architectural decisions
3. Speed Without Direction Is Waste
Soderstrom's comment about consumers being the limiting factor is telling. When you can produce software faster than users can absorb changes, the bottleneck shifts from engineering to product strategy. Small businesses actually have an advantage here -- you are closer to your customers, so you can iterate based on real feedback instead of shipping features into the void.
4. The Cost Equation Is Shifting
When AI writes the code, the value of a developer shifts from keystrokes to judgment. Senior engineers who understand your business domain, your architecture, and your customers become more valuable, not less. But you might need fewer of them, and they will spend their time differently. Baker McKenzie's recent restructuring showed a similar pattern in professional services -- fewer people doing more work, with AI handling the production layer.
The Bigger Picture
Spotify is not an outlier. This is happening at companies of every size, across every industry. Google recently launched Gemini 3.1 Pro, its most capable coding model yet, with SWE-Bench scores that would have been considered science fiction two years ago. OpenAI's Codex has evolved into a full development environment. Every major AI lab is investing heavily in code generation because they know the developer workflow is one of the first places where AI delivers measurable, undeniable ROI.
The question is no longer whether AI will write your code. It already does. The question is whether your team is set up to supervise it well.
What to Do Next
If you have not started integrating AI coding tools into your development workflow, the Spotify earnings call should be your wake-up call. Here is where to start:
- Pick one tool and try it. Claude Code, GitHub Copilot, or Cursor are all solid starting points. Give your senior developers a month with one of them.
- Measure what changes. Track pull request volume, review time, and bug rates. You need data, not vibes.
- Redesign your review process. AI-generated code is different from human-written code. Your review process should account for that.
- Talk to your team. AI fatigue is real. If your developers are burning out on code review, the tool is not working -- your process is.
The shift Spotify described is not coming. It happened in December. The only question is how quickly you catch up.
Need help integrating AI development tools into your workflow? Contact BaristaLabs -- we help small and mid-size businesses adopt AI the right way, without the burnout.