1. Slack, the popular workplace communication platform, has been utilizing files from user chats to train its machine learning models. The article reveals that Slack has been collecting and analyzing the content of these files, including text, images, and other media, to improve its natural language processing and other AI-powered features.

2. This practice has raised concerns about privacy and data usage, as users may not have been fully aware that their shared files were being used for this purpose. The article notes that Slack's terms of service do mention the potential use of user data for product improvement, but some argue that the specific details of this practice were not sufficiently disclosed.

3. The article suggests that Slack's use of user data for machine learning model training could be a common practice among tech companies, as they seek to leverage user-generated content to enhance their products and services. However, this raises questions about the transparency and consent involved in such data usage, and the potential implications for user privacy and trust.