Slack AI
Salesforce · Productivity · US · slack.com/features/ai
Last verified 2026-09-16
Entry
services/slack-ai.yamlTraining on your data (default tier)
Customer Data not used to train generative AI models without affirmative opt-in; models hosted inside Slack's trust boundary and do not retain requests; classic ML models use Customer Data unless the workspace opts out
Retention
Language models do not retain Customer Data after a request; message retention follows the workspace's retention settings
Data residency
Models run on Slack's cloud infrastructure within its trust boundary; region follows the workspace's data residency setting
Vendor data-handling terms
Training on your data (default tier)
Customer Data not used to train generative AI models without affirmative opt-in; models hosted inside Slack's trust boundary and do not retain requests; classic ML models use Customer Data unless the workspace opts out
“Slack will not use Customer Data to train generative AI models unless Customer provides affirmative opt-in consent. AI in Slack uses off-the-shelf LLMs where the models are not updated by and don't in other ways retain Customer Data after a request to them. AI in Slack hosts these models on its cloud provider (e.g., AWS) infrastructure, Customer Data stays in Slack's trust boundary, and the providers of the LLM don't have access to Customer Data. / If you want to exclude your Customer Data from helping train Slack global models, you can opt out.”
Retention
Language models do not retain Customer Data after a request; message retention follows the workspace's retention settings
“AI in Slack uses off-the-shelf LLMs where the models are not updated by and don't in other ways retain Customer Data after a request to them.”
Data residency
Models run on Slack's cloud infrastructure within its trust boundary; region follows the workspace's data residency setting
“AI in Slack hosts these models on its cloud provider (e.g., AWS) infrastructure, Customer Data stays in Slack's trust boundary, and the providers of the LLM don't have access to Customer Data.”
Domains and endpoints
observed means seen in Unseen deployments; vendor-documented means listed by the vendor. Vendors do not publish complete lists.
| Host | Role | Source |
|---|---|---|
| slack.com | app | vendor-documented |
| app.slack.com | app | vendor-documented |
| api.slack.com | api | vendor-documented |
Assessment
Workspace add-on: low
Reasoning: The AI summaries and search built into Slack, bought per workspace. Generative models run inside Slack's own infrastructure with no provider access and no training without opt-in. The separate opt-out for Slack's classic global ML models is the one admins tend to miss.
Flags: Models hosted inside Slack, no provider access · Classic ML training opt-out per workspace · Assessed 2026-09-16. The assessment is Unseen's; the terms above are the vendor's.
Changelog
- 2026-09-16Documented. Training, retention and hosting terms verified against Slack's privacy principles for search, learning and AI.
- 2026-09-16Added with vendor-documented domains.
Corrections
Pull request on GitHub, or the form below. Changes are reviewed and recorded in the changelog.