That's simple. If it is a model that was trained with only that person's data and is therefore connected to it, it is PII (personally identifiable data) and must be deleted.
If the model aggregates the data of multiple people and is disconnected from the individual it does not constitute PII and can be kept.
Most projects in the blockchain space focus on working towards an ICO to finance the operation. For many projects that makes sense. Sometimes however gearing the concept towards an incentive structure suitable for an ICO leads to non-optimal outcomes.
It also leads to a lack of generic infrastructure that is designed to be soundly engineered and easily reuseable.
What would you like to tackle during such a workshop?