Slack AI Training with Customer Data

This "ai" scam going on now is the ultimate convoluted process to hide sooo much tomfuckery: theres no such thing as copyright anymore! this isn't stealing anything, its transforming it! you must opt out before we train our model on the entire internet! (and we still won't spits in our face) this isn't going to reduce any jobs at all! (every company on earth fires 15% of everyone immediately) you must return to office immediately or be fired! (so we get more car data teehee) this one weird trick will turn you into the ultimate productive programmer! (but we will be selling it to individuals not really making profitable products with it ourselves)

and finally the most aggregious and dangerous: censorship at the lowest level of information before it can ever get anywhere near peoples fingertips or eyeballs.

> "These types of thoughtful personalizations and improvements are only possible if we study and understand how our users interact with Slack."

LOL

The OP privacy policy explicitly states that autocompletion algorithms are part of the scope. "Our algorithm that picks from potential suggestions is trained globally on previously suggested and accepted completions."

And this can leak: for instance, typing "a good business partner for foobars is" might not send that text upstream per se, but would be consulting a local model whose training data would have contained conversations that other Slack users are having about brands that provide foobars. How can Slack guarantee that the model won't incorporate proprietary insights on sourcing the best foobar producers into its choice of the next token? And sure, one could build an adversarial model that attempts to minimize this kind of leakage, but is Slack incentivized to create such a thing vs. just building an optimal autocomplete as quickly as possible?

Even if it were just creating classifiers, similar leakages could occur there, albeit requiring more effort and time from attackers to extract actionable data.

I can't blame Slack for wanting to improve their product, but I'd also encourage any users with proprietary conversations to encourage their admins to opt out as soon as possible.

> Why would anyone not opt-out?

Because you might actually want to have the best possible global models ? Think of "not opting out" as "helping them build a better product". You are already paying for that product, if there is anything you can do, for free and without any additional time investment on your side that makes their next release better, why not do it ?

You gain a better product for the same price, they get a better product to sell. It might look like they get more than you do in the trade, and that's probably true; but just because they gain more does not mean you lose. A "win less / win more" situation is still a win-win. (It's even a win-win-win if you take into account all the other users of the platform).

Of course, if you value the privacy of these data a lot, and if you believe that by allowing them to train on them it is actually going to risk exposing private info, the story changes. But then you have an option to say stop. It's up to you to measure how much you value "getting a better product" vs "estimated risk of exposing some information considered private". Some will err on one side, some on the other.