OpenAI is testing a new ChatGPT feature called “Writing Style” that can learn from examples in a user’s connected apps, according to BleepingComputer. The outlet reports that OpenAI confirmed it is testing the feature, but broad availability timing is not known. The feature is available to a small group of users, BleepingComputer reports. An onboarding screen described by the outlet says ChatGPT will be able to write in the user’s voice by referencing examples from connected apps. The setup shown by BleepingComputer groups examples into three categories: Messaging, Documents and Email. The services listed include Slack for messaging, Google Drive and Notion for documents, and Gmail for email. The practical change is in how ChatGPT may gather style context. Instead of asking users to repeatedly describe their preferred tone or paste samples into a chat, the test would let ChatGPT reference writing that already exists in connected services, according to the report. BleepingComputer frames the approach as distinct from Anthropic’s earlier Styles feature, which let users upload or paste samples and generate a custom style based on those examples. OpenAI’s test, as reported, would rely on material already inside connected apps, potentially giving ChatGPT examples across different formats such as short messages, emails and longer documents. The report does not establish when the feature will reach more users, what controls will be available beyond the onboarding flow described, or how broadly ChatGPT will use examples from each connected service. For now, the confirmed fact is narrower: OpenAI is testing Writing Style, and the visible configuration points to Slack, Google Drive, Notion and Gmail as supported sources in the test. Who benefits: Users who frequently draft emails, messages or documents in ChatGPT could benefit if the feature reduces the need to provide style instructions repeatedly. Teams already using Slack, Gmail, Google Drive or Notion are the clearest fit based on the reported integrations. Who's exposed: The main exposure is for users and organizations connecting personal or work apps to ChatGPT without clear internal rules for what content should be referenced. The available report does not provide enough detail to assess enterprise controls or data-handling limits.