Users of Salesforce’s popular Slack messenger noticed that the service uses messages and files to train AI. Information about this is contained in a confusing privacy policy — and users have to write to the company themselves.
How to tells us ArsTechnica, it’s all begun from a thread on the Hacker News forumwhere users accused Slack of collecting data to train AI without warning anyone.
Meanwhileon the Slack AI pageHe said: «Work without worry. Your data — is your data. We don’t use it to train Slack AI».
To opt out, users have to send an email to the company themselves.
But the main problem remains — users did not explicitly consent to the exchange of data for use in AI training.
Due to this inconsistency, users have called on Slack to update its privacy guidelines to make it clear how data is used for any future AI updates.
Slack and Salesforce responded to the online discussion with rather vague posts, apologies, and promises to update their privacy policies.
«We don’t create or train these models so that they can learn, memorize, or reproduce a certain part of customer data,» the company representative explained.
The company also emphasized that it does not use data for Slack AI.
Customers’ data is used specifically for training «global models» — we are talking about recommendations for channels, emojis, and search results.
Nevertheless, the European Union’s General Data Protection Regulation (GDPR) requires informed and specific consent before companies can collect data. That is, consent cannot be given by default, it must be specifically recorded.
Slack has stated that it is committed to adhere to GDPR by promising to «update our product features and contractual obligations».