Microsoft's red team has monitored AI since 2018. Here are five big insights

Microsoft is taking measures to ensure AI safety in its models and wants to set an example for others.