In a recent development, popular chat platform Slack, owned by Salesforce, has found itself at the center of controversy surrounding its use of user data for training AI services. The issue came to light when a user posted about it on a popular community site for developers, sparking a viral discussion.

Slack, like many other tech companies, leverages user data to train its AI services. However, it has emerged that users are automatically opted in to this data usage, with the option to opt out only available upon emailing the company. The privacy policy, which contains these terms, was described as outdated and confusing, leaving many users unaware of Slack's AI training practices.

The Hacker News thread discussing Slack's AI services brought up several questions and concerns, particularly regarding the lack of clear information about the company's "Slack AI" product and its relationship to the privacy policy. Users were left puzzled by the references to "global models" and "AI models," while wondering where exactly Slack applies its AI privacy principles.

Slack has clarified that it uses customer data specifically for training "global models" that power features such as channel and emoji recommendations and search results. However, the wider scope of its AI model training remains unclear in the current policy. Despite the ability to opt out of data training, users still benefit from Slack's "globally trained AI/ML models," leaving them questioning the necessity of using customer data for features like emoji recommendations.

The company has stated that it does not use customer data to train Slack AI, a separate add-on that utilizes large language models (LLMs). Slack AI uses LLMs hosted within the company's AWS infrastructure, ensuring customer data remains in-house and exclusively for the organization's use.

The confusion surrounding Slack's AI practices has prompted discussions on the need for clearer privacy principles. One Slack engineer concurred that the company should update its privacy policy to reflect how it relates to Slack AI, admitting that the terms were written before the introduction of this product.

The controversy at Slack serves as a reminder of the importance of user privacy in AI development. Companies must clearly outline their data usage policies to ensure transparency and avoid surprises for their users.

As the backlash continues to grow, it is expected that Slack will address the concerns and provide further clarity on its AI practices in the near future. This incident highlights the need for companies to prioritize user privacy and keep their terms of service updated to align with evolving AI technologies.

If you have a news tip, you can contact Ingrid securely via Signal at ingrid.101 or submit your tip here. Additionally, starting from June 5, we will be launching an AI newsletter. Sign up here to receive it directly in your inbox.