Google warns own staff about entering data into, using code from, chatbots

Re:Google warns own staff about entering data into, using co... by filoleg on Thursday June 15, 2023 @12:57PM (#36343411)

No, because those products are controlled by Google aka their employer, so it is on the employer to make sure no confidential data leaks to the outside from its own products. And that is both well within their control and responsibility, and I am fairly certain they accept this as a fact.

But there is exactly zero control Google has over the confidential data entered in tools and forms that don't belong to them.

[ Parent | on HN ]

Re:Google warns own staff about entering data into, using co... by bluefishinit on Thursday June 15, 2023 @01:10PM (#36343624)

I meant from the perspective of a company (not Google) using Google's service. Essentially, Google is making the argument that their own saas products are untrustworthy.

[ Parent | on HN ]

Re:Google warns own staff about entering data into, using co... by vkou on Thursday June 15, 2023 @01:25PM (#36343822)

Speaking in general terms, when you buy a corporate subscription to those kinds of SAAS products, they will include contractual provisions that bar your vendor from looking at your data.

Your use of ChatGPT is not protected by any such provisions.

[ Parent | on HN ]

Re:Google warns own staff about entering data into, using co... by cududa on Thursday June 15, 2023 @03:35PM (#36345388)

You're making a LOT of logical leaps here - did you even bother to read the article? How in the world do you go from "Don't use confidential information in public chat bot instances" to "Google Docs is used in training AI models and therefore is untrustworthy"?

[ Parent | on HN ]