Seven good points to consider when making an AI policy:

• Purpose and scope: how is it being used?
• Data governance: collection, storage, use, anonymisation
• Model training: especially re free versions and what information you are giving models
• Transparency and accountability: explainable decision-making, catch biases/errors, human in the loop
• Informed consent and privacy: Clear to supporters how their data is used, if sending sensitive data to AI.
• Bias and fairness: monitoring processes, regular assessment.
• Training and capacity building:
• Monitoring and evaluation: clear schedule to check AI systems