AI use
AI use and ethics statement
Sendoris uses AI as a tool, never as an unsupervised decision maker. A named person is accountable for every recommendation we give and every system we build.
Last updated
How we use AI in our own work
We use AI to research, summarize and draft. Everything that reaches a client is reviewed by a consultant who is accountable for it.
No article on this website is published automatically. Content assisted by AI is read, corrected and approved by a person before it goes live.
How we design AI for clients
We keep a human in the loop wherever a mistake would be expensive, irreversible or unfair to someone.
We prefer the simplest approach that works. If ordinary logic solves the problem, we do not add a model to it.
We document what a system does, what data it uses and how it can be switched off.
Data and confidentiality
Client data is used for that client's work only. We discuss data handling, retention and location before any system touches real information.
What we will not do
We do not build systems whose purpose is to mislead people, to hide that they are automated where that matters, or to make consequential decisions about individuals without human review.