Slack has clarified its AI and ML policy, providing more information regarding how it does and does not use customer data.

Slack ruffled feathers last week when it unveiled terms that seemed to indicate the company was using customer data to train AI/ML models. The relevant text is below:

Machine Learning (ML) and Artificial Intelligence (AI) are useful tools that we use in limited ways to enhance our product mission. To develop AI/ML models, our systems analyze Customer Data (e.g. messages, content, and files) submitted to Slack as well as Other Information (including usage information) as defined in our Privacy Policy and in your customer agreement.

To make matters worse, the company requires Workspace Owners to manually email the company to opt out, rather than providing a simple option in the app or on the website.

The company has since clarified its stance, emphasizing that it does not use customer data for developing or training large language models (LLMs). A Salesforce spokesperson provided the following information to WPN:

• Slack has industry-standard platform-level machine learning models to make the product experience better for customers, like channel and emoji recommendations and search results. These models do not access original message content in DMs, private channels, or public channels to make these suggestions. And we do not build or train these models in such a way that they could learn, memorize, or be able to reproduce customer data.
• We do not develop LLMs or other generative models using customer data.
• Slack uses generative AI in its Slack AI product offering, leveraging third-party LLMs. No customer data is used to train third-party LLMs.
• Slack AI uses off-the-shelf LLMs where the models don’t retain customer data. Additionally, because Slack AI hosts these models on its own AWS infrastructure, customer data never leaves Slack’s trust boundary, and the providers of the LLM never have any access to the customer data.

Note that we have not changed our policies or practices – this is simply an update to the language to make it more clear.

In an accompanying blog post, the company says that it does not leak data across workspaces, meaning it does “not build or train ML models in a way that allows them to learn, memorize, or reproduce customer data.” The company also says that “ML models never directly access the content of messages or files.” Instead, the company uses numerical features, such as message timestamps, the number of interactions between users, and the number of overlapping words in the channel names a user is member of, to make relevant suggestions.

The company emphasized its industry-standard privacy measures designed to protect customer data.

Slack uses industry-standard, privacy-protective machine-learning techniques for things like channel and emoji recommendations and search results. We do not build or train these models in such a way that they could learn, memorize, or be able to reproduce any customer data of any kind. While customers can opt-out, these models make the product experience better for users without the risk of their data ever being shared. Slack’s traditional ML models use de-identified, aggregate data and do not access message content in DMs, private channels, or public channels.

Unfortunately, Slack still requires an email be sent to the company to opt out, making the case that a mass opt-out from its users would make the overall experience worse.

Customers can email Slack to opt out of training non-generative ML models. Once that happens, all data associated with your workspace will be used to improve the experience in your own workspace. You will still enjoy all the benefits of our globally trained ML models without contributing to the underlying models. No product features will be turned off, although certain places in the product where users could have previously given feedback will be removed. The global models will likely perform slightly worse on your team, since any distinct patterns in your usage will no longer be optimized for.

In other words, any single customer opting out should feel minimal impact, but the greater number of customers who opt out, the worse these types of models tend to perform overall.

While the company makes a valid point, it should still provide an easier way for customers to opt out, with a clearly labeled message explaining the downsides of doing so. Requiring users to send an opt-out email is an unnecessary roadblock the company is artificially creating in an effort to dissuade users from taking advantage of the option.