Ethics and responsible AI
Human judgement, fairness and accountability belong at the centre of technology decisions.
Start with a genuine need
Use the simplest suitable approach. The right answer may be a person, an existing tool, a template or conventional automation rather than AI.
Keep people responsible
AI output needs appropriate human review. Training can reduce avoidable risk but cannot guarantee accuracy, fairness, compliance or a business outcome.
Respect privacy and fairness
Use approved tools and authorised information only. Do not enter confidential information into demonstrations. Check outputs for exclusion, bias and harm, especially when a decision affects people.
Measure responsibly
Use transparent, proportionate and preferably aggregate measures. Activity data must not become an automatic employee performance score or hidden monitoring tool.
Raise a concern
Use our contact page or write to the registered office. Privacy concerns can also be raised through the route in our privacy notice.