The State of AI in Platform Engineering has launched! You can download it now.

240+ platform engineers surveyed, dozens of deep interviews, 100s of hours of reading and watching, and a staggering shift in how we’re all using (and struggling with) AI.

What did we learn? 89% of teams are using AI but the vast majority of teams are seeing zero value. Those who are? It’s bc they have an effective Internal Developer Platform as a foundation for their AI goals.

That’s why this report aims to go further than just research, it’s also a playbook for how you can succeed in AI.

• Two worlds of AI in platforms: We kept seeing confusion between AI-powered platforms (using AI to enhance your developer platform i.e troubleshooting, natural language interfaces, policy enforcement) and platforms for AI (infra for ML engineers, scalable compute, model pipelines, secure data). Mixing the two leads to wasted sprints and mismatched goals
• The “AI implementation plateau”: Teams race in with gen code and docs, leadership gets hyped and sets unrealistic or unthought out goals, then everything stalls. Slow progress, scaling issues, security fears, hallucinations, and skill gaps hit hard. The small % of winners are the ones who blast through this plateau.
• Impact isn’t automatic: Code generation feels magical, but it doesn’t move organizational productivity alone. Real ROI comes from targeted improvements like automating error triage, smarter observability, safer pipelines not blanket AI mandates. Blanket AI mandates are as detrimental as blanket AI bans.

These are five principles to breaking through the plateau we’ve identified across top teams.

• AI amplifies your platform’s flaws, so fix your foundations first.
• Start with clear outcomes (I.e “Reduce error analysis time”, not something vague like “use more AI”).
• Treat AI features like a product
• Build pragmatic governance. Safe experimentation > blanket bans.
• Invest in your team’s skills so AI isn’t shadow IT

This report is a reality check. AI’s everywhere, but value only comes when strong platform foundations meet clear intent and thoughtful rollout. Across 40+ pages, it breaks down each of these bullet points to help ensure you are the 5% who see real AI value in the enterprise.

The opportunity has never been bigger for platform engineers.