Hacker News Reader

Comment by andyxor | Hacker News Reader

Google response (from Jan):

"Our security systems automatically lock an employee’s corporate account when they detect that the account is at risk of compromise due to credential problems or when an automated rule involving the handling of sensitive data has been triggered. In this instance, yesterday our systems detected that an account had exfiltrated thousands of files and shared them with multiple external accounts. We explained this to the employee earlier today. We are actively investigating this matter as part of standard procedures to gather additional details.”

https://venturebeat.com/2021/01/20/google-targets-ai-ethics-...

View on HN

That seems like a pretty serious accusation. Looks like relations between management and employees has seriously deteriorated.

A company with 100 000 people will have to quite a number of staff, literally daily, for all sorts of things, it doesn't remotely imply anything about the overall relationship between staff and management.

Even when CNN, MSNBC, Business Insider 'team up' to overrepresent an issue because 'some number of Google employees walked out for some reason' - it really doesn't help tell the story of 'relations' because it's hard to fathom what the rest of the 99% of Google staff think.

I meant to say relations between management and staff in the AI ethics division.

Always seems to be from the same divisions.

That said, anything that makes Google a less effective organization I support.

Funny how they co-announced monetary incentives around "diversity and inclusion", I expect nothing less from a giant evil ad company. A bunch of handwavy idpol stuff with the giant elephants in the room seems to be the status quo.

AI system fires AI ethics leader? Hmm this wasn't the machine uprising I was expecting.

how would this system detect sharing with external accounts? sounds like they put some words together to make it sound like it’s all standardized and automated and not result of human actions

You are wrong.

All outgoing email is being scanned for sensitive information in plain text or attachments. This is standard practice in every corporation and something you are being warned about when being on-boarded as a new-joiner.

why would she email it from her corp account to various accounts instead of emailing it to her personal account or just uploading it directly to some personal drive account

> all outgoing email

And of course this same applies to file uploads. Again, this is standard practice in any company with IT security.

> an account had exfiltrated thousands of files and shared them with multiple external accounts

I'm just questioning why she would email to multiple external accounts directly from her corp email instead of saving to 1 personal account and then sharing from there, undetectable, many times. And yes, they would still have detected the 1 email but I'm questioning the accuracy of this explanation; sounds partly made up to make it sound more serious.

Multiple external accounts could mean her lawyer's account and her own. Or 2 lawyers' accounts. The latter would help show she didn't distribute the files to anyone else.

> sounds like they put some words together to make it sound like it’s all standardized and automated and not result of human actions

Part of my job is administering a system for detecting unauthorized sharing, exporting, or printing. It has thresholds and a rules-based engine for specifying what behavior should trigger alerts. Google's system is probably custom but there are definitely products on the market for this sort of thing.

Reply on news.ycombinator.com