Databricks has made its Genie app for Slack generally available, bringing company-data questions and some cross-app actions into chat. The useful detail is the split between what Genie can access and who may see its answers: access follows the asker’s permissions, but channel replies are visible to the whole channel by default.
Databricks Watch analysis
What happened
Staff can message Genie directly or use Genie in a Slack channel. It replies in the same thread, can cite sources including Databricks data, Google Drive, Confluence and Jira, and can analyse uploaded files, make charts and send scheduled reports. Databricks also says Genie can update documents, comment on tickets and change records in connected tools. The person who requested a change must approve it first.
That permission check does not make channel replies private. Other channel members can read an answer even if they do not have access to the underlying data. Admins can set all answers to private. Databricks says questions are processed in the customer’s workspace and the service relaying messages between Slack and Databricks does not store them. It also warns that Genie can give wrong answers, so users should check before acting. The Next Web’s report details the launch.
Why it matters
Genie is more than a chat window for looking up company figures: it can propose changes across connected tools, with approval required before those changes happen. That could save staff a trip between apps, while putting the approval step where it belongs, with the person asking.
But permission to retrieve data and permission to disclose an answer are not quite the same thing. In a busy channel, an answer visible to everyone could reveal information to colleagues who could not open the source themselves. Admins have a privacy setting; organisations will need to decide whether the default suits the way their channels are used.
Our read
The useful part is the combination of cited answers and approval before changes, not the familiar promise that AI will make work feel frictionless. Try it on a low-stakes task first, check the sources and be deliberate about whether answers belong in a channel or a private message. A permission check is welcome; it is not a substitute for thinking about the audience.
What to watch
- Whether admins can set privacy defaults that match how different teams use Slack.
- How well Genie’s citations help staff spot incorrect answers.
- Which connected-tool changes it supports in practice, and how clearly approvals are recorded.
Discussion spark: Should workplace AI answers be visible to everyone in a channel by default, or should they be private unless the asker chooses to share them?
Sources and evidence
- Source update (12 November 2026, 16:55 UTC)
Independent WittyWires tracker for public updates about Databricks. Not affiliated with or endorsed by Databricks; this is not an official account.