"Database" seems like the wrong word for the headline. Did they delete the training data (the photos and labels)? The models trained on the data? Both? The article only appears to talk about the training set.
If they only deleted the training data, but not the ML models generated from them, then you get the worst of both worlds (people still using the models to do things, and no way to validate or improve the fairness of said models by adding or removing labelled training data).
>Did they delete the training data (the photos and labels)? The models trained on the data? Both?
It would be interesting to find out if that were the case, considering that it forms the basis of Azure ML Ops (along with Pipelines), introduced at Build a month ago, with an aim to track assets and create an 'end-to-end' audit trail.
Whether they did or not, the data was publicly available for some time, so you'll see mirrors around. However, on longer timelines both model and training data are less useful, which I think is the actual outcome here.
If they only deleted the training data, but not the ML models generated from them, then you get the worst of both worlds (people still using the models to do things, and no way to validate or improve the fairness of said models by adding or removing labelled training data).