Not to be glib, but this why we built Tonic Textual (www.tonic.ai/textual). It’s both very challenging and very important to protect data in training workflows. We designed Textual to make it easy to both redact sensitive data and replace it with contextually relevant synthetic data.
Slack AI Training with Customer Data
11–20 of 426 posts
Re: Slack AI Training with Customer Data
#12I'm confused about this statement: "When developing AI/ML models or otherwise analyzing Customer Data, Slack can’t access the underlying content. We have various technical measures preventing this from occurring" "Can't" is a strong word. I'm curious how an AI model could access data, but Slack, Inc itself couldn't. I suspect they mean "doesn't" instead of "can't", unless I'm missing something.
From their white paper linked in the same comment
> Provisioning To minimize the risk of data exposure, Slack adheres to the principles of least privilege and role-based permissions when provisioning access—workers are only authorized to access data that they reasonably must handle in order to fulfill their current job responsibilities. All production access is reviewed at least quarterly.
so... seems like they very clearly can.
Re: Slack AI Training with Customer Data
#13Eugh. Has anyone compiled a list of companies that do this, so I can avoid them? If anyone knows of other companies training on customer data without an easy highly visible toggle opt out, please comment them below.
We can fight back by not posting anything useful or accurate to the internet until there are protections in place and each person gets to decide how their data is used and whether they are compensated for it.
[flagged]
Re: Slack AI Training with Customer Data
#14Eugh. Has anyone compiled a list of companies that do this, so I can avoid them? If anyone knows of other companies training on customer data without an easy highly visible toggle opt out, please comment them below.
[flagged]
Re: Slack AI Training with Customer Data
#15Eugh. Has anyone compiled a list of companies that do this, so I can avoid them? If anyone knows of other companies training on customer data without an easy highly visible toggle opt out, please comment them below.
Synology updated this policy back in March (Happened to be a Friday afternoon).
Services Data Collection Disclosure
"Synology only uses the information we obtain from technical support requests to resolve your issue. After removing your personal information, we may use some of the technical details to generate bug reports if the problem was previously unknown to implement a solution for our products."
"Synology utilizes the information gathered through technical support requests exclusively for issue resolution purposes. Following the removal of personal data, certain technical details may be utilized for generating bug reports, especially for previously unidentified problems, aimed at implementing solutions for our product line. Additionally, Synology may transmit anonymized technical information to Microsoft Azure and leverage its OpenAI services to enhance the overall technical support experience. Synology will ensure that personally identifiable information, such as names, phone numbers, addresses, email addresses, IP addresses and product serial numbers, is excluded from this process."
I used to just delete privacy policy update emails and the like but now I make a habit of going in to diff them to see if these have been slipped in.
Re: Slack AI Training with Customer Data
#16> Data will not leak across workspaces.
> If you want to exclude your Customer Data from helping train Slack global models, you can opt out.
I don't understand how both these statements can be true. If they are using your data to train models used across workspaces then it WILL leak. If they aren't then why do they need an opt out?
Edit: reading through the examples of AI use at the bottom of the page (search results, emoji suggestions, autocomplete), my guess is this policy was put in place a decade ago and doesn't have anything to do with LLMs.
Another edit: From https://slack.com/help/articles/28310650165907-Security-for-...
> Customer data is never used to train large language models (LLMs).
So yeah, sounds like a nothingburger.
Re: Slack AI Training with Customer Data
#17Eugh. Has anyone compiled a list of companies that do this, so I can avoid them? If anyone knows of other companies training on customer data without an easy highly visible toggle opt out, please comment them below.
It would be easier to compile a list of companies that don't do this.
The list:
Re: Slack AI Training with Customer Data
#18> Data will not leak across workspaces. > If you want to exclude your Customer Data from helping train Slack global models, you can opt out. I don't understand how both these statements can be true. If they are using your data to train models used across workspaces then it WILL leak. If they aren't then why do they need an opt out? Edit: reading through the examples of AI use at the bottom of the page (search results…
There are no data leaks in Ba Sing Se.
Re: Slack AI Training with Customer Data
#19> Data will not leak across workspaces. > If you want to exclude your Customer Data from helping train Slack global models, you can opt out. I don't understand how both these statements can be true. If they are using your data to train models used across workspaces then it WILL leak. If they aren't then why do they need an opt out? Edit: reading through the examples of AI use at the bottom of the page (search results…
They're saying they won't train generative models that will literally regurgitate your text, my guess is classifiers are fair game in their interpretation
Re: Slack AI Training with Customer Data
#20I'm confused about this statement: "When developing AI/ML models or otherwise analyzing Customer Data, Slack can’t access the underlying content. We have various technical measures preventing this from occurring" "Can't" is a strong word. I'm curious how an AI model could access data, but Slack, Inc itself couldn't. I suspect they mean "doesn't" instead of "can't", unless I'm missing something.
I also find the word "Slack" in that interesting. I assume they mean "employees of Slack", but the word "Slack" obviously means all the company's assets and agents, systems, computers, servers, AI models, etc.
I would find even a statement from Signal like "we can't access our users content" to be tenuous and overly-optimistic. Like, when I heard the word "can't" my brain goes to: there is nothing anyone in the company could do, within the bounds of the law, to do this. Employees at Slack could turn off the technical measures preventing this from occurring. Employees at Signal could push an app update which side-channels all messages through to a different server, unencrypted.
Better phrasing is "Employees of Slack will not access the underlying content".