While it does seem to have vanished, this has some rather pronounced data security concerns for my org. What information within our instance will this tool be capable of accessing, how does it process that data, is it being used to train a public LLM? The option to 'summarize this document' implies ...