AI use and governance policy
Today
OBMr does not use artificial intelligence or machine learning to process customer data. No information in OBMr is sent to an AI model provider, and none is used to train a model. The trends and summaries in the app are ordinary arithmetic you can check by hand: means, counts and a least-squares slope.
If we add AI features
Any AI feature will follow these rules before it is switched on:
- Off until the organization turns it on. Each customer decides, and can turn it off again.
- A person decides. AI may draft, summarize or suggest. It will not hire, discipline, rate, rank, pay, promote or end anyone's employment, and it will not make any decision with a legal or similarly significant effect on a person.
- Labelled. Anything an AI produced is marked as such where people see it, with the information it was based on.
- No sensitive inferences. No emotion recognition, no inferring health, disability, or other protected characteristics, and no profiling based on them.
- Assessed first. We will document the purpose, the data used, the risks to employees, and the legal classification. Employment uses are treated as potentially high-risk, including under the EU AI Act where it applies.
- Contracted providers only. A model provider is added only with a written agreement that forbids training on customer data and limits retention, and only after it appears on our subprocessors page with notice to customers.
- Logged. Every AI request is recorded with who made it and why.
Your organization's own AI tools
If your organization uses other AI tools, do not paste OBMr records into them unless your organization has approved that tool for employee personal information.
Questions: privacy@obmr.app.