Let’s be real, folks — few reports in our industry carry the weight and staying power of the DORA report. For over a decade now, the “Accelerate State of DevOps” and its successors have shaped how we think about performance, culture and the DNA of high-functioning software teams. The famous four key metrics — deployment frequency, lead time for changes, mean time to restore, and change failure rate — are practically the commandments of DevOps. They gave us a language, a framework, and frankly, a way to separate myth from measurable progress.
I can tell you firsthand: I’ve seen DORA findings cited in boardrooms, in funding decks, and at just about every major DevOps event since 2014. When the report first showed that elite performers could deploy hundreds of times per day with lower failure rates than their peers, it blew the doors off the old “faster means sloppier” mindset. That single data-driven insight legitimized what a lot of us had been preaching but couldn’t prove. DORA turned DevOps from a movement into a management conversation.
Which brings us to 2025. The newest DORA report — the “State of AI-Assisted Software Development” — lands at a time when AI is eating everything from code generation to documentation to operations. And just like those early DORA reports reframed speed versus stability, this one is reframing what AI is actually doing to our software delivery pipelines. Spoiler alert: It’s not as simple as “AI makes everything better.”
Adoption Outpaces Trust
Here’s the headline: A full 90% of respondents say they’re now using AI in their work, and most believe it’s boosting productivity. That’s near-universal adoption, folks. But here’s the rub — about a third admit they don’t trust AI-generated code.
That’s the kind of tension that defines inflection points. Remember when we first started adopting CI/CD pipelines? Everyone set them up because it was the thing to do, but very few trusted them enough to deploy to production on a Friday afternoon. Same story here. AI is table stakes already, but confidence hasn’t caught up with usage.
And if your engineers don’t trust the code, what do you really have? You’ve got more friction, not less.
Faster, Yes. Safer? Not So Much
Now here’s the counterintuitive part. For the first time, DORA shows AI adoption is linked to higher throughput. That’s right — teams using AI are moving work through the system faster than those who aren’t.
But before you pop the champagne, look at the other half of the finding: Instability is still higher in AI-heavy teams. Faster, yes. Safer? Not so much.
If you’ve been around the block, this won’t shock you. We saw the same thing in the early days of automation — speed without discipline just meant you hit the wall quicker. To borrow a metaphor, AI is rocket fuel. But without a guidance system, you’re just accelerating toward the cliff.
Platforms Make or Break the AI Payoff
Here’s a finding that shouldn’t surprise anyone who’s been paying attention: Platform quality is the difference-maker. A full 90% of organizations now have internal platforms, and three-quarters have dedicated platform teams. But the data shows a stark divide. High-quality platforms amplify AI’s benefits across the board. Low-quality platforms? AI impact is negligible.
Truth is, this aligns perfectly with the rise of platform engineering as a discipline. We’ve been saying it for years — platforms are the backbone of modern DevOps. Now DORA is proving it with data. If your internal platform stinks, all the AI in the world won’t save you. In fact, it’ll probably just make the stink spread faster.
VSM: The Unsung Hero
Another gem buried in the report is the role of value stream management. AI tends to deliver “local optimizations” — an engineer codes faster, a test suite runs quicker — but without VSM, those wins don’t always roll up into business outcomes. With VSM in place, AI-driven productivity gains translate into measurable improvements at the team and product level.
That, to me, is vintage DORA. Remember when they proved that culture — psychological safety, autonomy, collaboration — wasn’t just a warm fuzzy HR concept but directly correlated with elite performance? Same here. VSM turns AI from a toy into a force multiplier.
The User-Centric Reality Check
One more finding worth calling out: Teams that adopt AI without a strong user focus actually see performance harm. When you center the user, AI’s impact is strongly positive. When you don’t, it can backfire.
Think about that. All the talk about AI copilots and generative magic means nothing if the work isn’t aligned with what the end user actually needs. It’s a classic reminder that DevOps has always been about delivering value, not just velocity.
So What Does it all Mean?
Pulling back, here’s how I see it:
- The trust gap is real. AI may be everywhere, but confidence will take years of scars and stories to catch up.
- Speed without stability is the oldest tradeoff in software, and AI hasn’t rewritten that law yet.
- Platforms and VSM aren’t optional—they’re the levers that turn AI adoption into real organizational performance.
- User focus matters more than ever. AI without empathy just means scaling the wrong work faster.
If this all sounds familiar, it should. The early DORA reports shattered the myth that speed and stability were opposites. The 2025 report is showing us that AI doesn’t change the fundamentals — it just raises the stakes.
What You Should Do Next
So what’s a DevOps leader supposed to do with all this? Here’s my playbook:
- Get your platform house in order. Don’t bolt AI onto a shaky foundation.
- Double down on VSM. Make sure productivity gains aren’t just treadmill miles.
- Define trust boundaries. Engineers should know when AI output is acceptable and when it needs review.
- Invest in rollback and safety nets. The faster you go, the better your brakes need to be.
- Keep the user at the center. AI is only as good as the product direction it’s pointed toward.
Shimmy’s Take
At the end of the day, DORA has always been the sober voice in a hype-driven industry. Back in 2014, it told us faster could be safer. In 2025, it’s telling us AI can be faster, but not necessarily safer — at least not yet.
If you’ve got your fundamentals in place — platforms, pipelines, VSM and a relentless focus on users — AI can give you a serious edge. But if you’re hoping AI will fix broken culture or bad process, I’ve got news for you. I’ve seen this movie before, and it doesn’t end well.
