JPMorganChase expands Claude Code to more employees across the firm

Does this mean just Claude Code runs on JPMorgan Chase infra? How does a firm like JPMorgan Chase justify sending all its sensitive company information to be collected by Anthropic?

When some employee copy pastes a conversation from a work chat, where is the log of that stored? My understanding it is sent to Anthropic who stores the data and "may" delete it after 30 days unless Zero Data Retention is agreed on and setup?

I guess surprises me that a security conscious organization like JPMorgan wouldn't look towards on premises open weight models where they are certain of the security, but maybe I'm too cynical