Anthropic has launched Claude Tag, a continuously active AI assistant integrated directly into Slack. The tool retains context and can independently act within team conversations.
Anthropic has introduced a new service called Claude Tag. The feature is designed as a permanently active AI assistant embedded directly into the communication platform Slack, where it operates as a virtual team member. The feature will be available to enterprise customers on Claude Enterprise and Claude Team plans as a beta version starting in June 2026. Users can bring the AI into conversations by mentioning its profile name to request analyses or assign tasks.
Compared with previous integrations, the system comes with persistent memory. The company explained:
“By having Claude follow the channel, it learns more and more about the work.”
Anthropic
If enabled by system administrators, the application can also combine information from other channels across the organization.
Administrative Controls and a Proactive Work Mode
Within a Slack channel, all team members interact with the same AI identity. This keeps task history visible to everyone involved, allowing conversations and workflows to continue seamlessly across different users. Administrators define which tools, data sources, and channels each instance can access. Memory content remains strictly limited to assigned areas, helping prevent unintended data transfers between departments such as legal and engineering.
When assigned tasks, the system breaks them down into individual steps and executes them using the available tools. Results are posted directly into the relevant Slack thread. In addition to direct mentions, the software includes an ambient mode. In this state, the AI can proactively join conversations to provide updates, highlight cross-organizational topics, or remind teams about unfinished tasks.
According to Anthropic, the interaction should feel like “working with a real coworker – one who can complete work visibly for everyone, with far more context and understanding than before.”
Competitive Race for Enterprise Context in the Cloud
Capturing and providing company-specific context is considered a key factor in deploying AI systems across organizations. Anthropic is not the only technology company pursuing this approach. Microsoft uses the Microsoft Graph framework together with Copilot and Work IQ to connect AI systems with enterprise knowledge. Data platforms such as Snowflake and Databricks position their technologies as foundations for providing organizational knowledge to autonomous agents. Glean is also developing a specialized intelligence layer designed to act as an intermediary between AI models and companies’ internal data environments.
(ll)