Principles
How we handle AI
AI is powerful, but not infallible. That's why we work with clear principles.
No black boxes. No uncontrolled automation. Instead: transparency, human oversight, and a healthy dose of pragmatism.
Our principles
A human always decides
AI advises. A human decides.
Whether it's code review, strategic analyses, or automated actions — there's always someone who has the final say. AI is an assistant, not a boss.
Transparent about what AI does
We tell you when AI is being used.
No hidden algorithms making decisions without your knowledge. When AI creates something, we say so. When AI draws a conclusion, we show what it's based on.
Your data stays yours
We don't train models on your data.
Your files, your conversations, your business information — none of it goes to third parties to make AI smarter. Privacy by design, not as an afterthought.
Honest about limitations
AI makes mistakes.
Sometimes hallucinations (making things up), sometimes simply wrong conclusions. We don't pretend AI is infallible. We build in checks, train people to stay critical, and are honest when something doesn't work as expected.
A tool, not a replacement
AI makes good people better. It doesn't replace them.
A developer with AI support is faster and more thorough than without. But it's still the developer who has the knowledge, makes the decisions, and is responsible for the result.
Continuously evaluate
Technology changes fast.
What works today might be outdated tomorrow. What's impossible today could be normal tomorrow. We regularly evaluate: does this still work? Can we now responsibly deploy this? Should we stop doing something?
Why this matters
You give us access to your code, your data, your business processes. That's trust. These principles are how we honor that trust. Not with nice words, but with concrete choices about what we do and don't do.