News

Claude Meetings Brings Connected Tools into Calls, With Different Defaults for Team and Enterprise

The Google Meet beta can create shared work during conversations, making tool permissions and the audience for its output central deployment questions.</

Anthropic is bringing Claude into video meetings as an active participant that can create documents and consult connected tools, extending the enterprise AI assistant’s role beyond recording what people said.

The company introduced Claude Meetings November 12. Its release notes describe a beta rollout for Team and Enterprise customers through Google Meet.

Participants can ask the assistant questions and have it build shared artifacts while the discussion continues. The enterprise question is how information retrieved with one person’s authorization becomes available within a group conversation.

Tool Permission and Audience Are Different Controls
According to the current help documentation, Claude uses each person’s connected tools with that person’s permission. It asks on first use, with approval applying going forward. Spoken requests currently require confirmation of the speaker’s identity.

The documentation distinguishes Meet chat, visible to everyone on the call, from the Claude meeting page’s internal chat. Participants should not assume that authorizing access to a source and approving the audience for an answer are the same decision.

For example, an employee may be entitled to retrieve an internal document while some meeting participants are not its intended recipients. That hypothetical illustrates why a collaboration feature needs controls over both retrieval and presentation.

External-Guest Defaults Depend on the Plan
The current help page says outside-guest meetings are allowed by default on Team plans and require owner approval on Enterprise plans. When such guests are allowed, Claude takes notes but does not reply in the call’s chat; guests do not receive the internal meeting page or recap email.

The launch article describes staying out of external meetings as a default without the same plan-level distinction. Administrators assessing the rollout should use the more specific current configuration documentation rather than infer settings from that general launch description.

The difference matters because enabling the feature does not necessarily establish the same audience boundary in every organization. It also means a meeting organizer needs to understand the applicable plan settings before using automatic joining.

Recording a Conversation Creates Another Information Asset
Anthropic says Claude saves a transcript, summary, and screen descriptions, rather than audio. The company describes a meeting page where invited members of the organization can access the resulting material and collaborate on artifacts.

Not storing audio does not mean that the conversation leaves no retained record. The transcript and generated documents can still contain sensitive information or conclusions that need correction.

Anthropic says the beta is unavailable to some organizations, including those using zero data retention, HIPAA-ready configurations, or customer-managed encryption keys. Those exclusions are limits on initial availability, not evidence that the feature meets every other organization’s requirements.

The immediate evaluation is consequently about more than note quality. Teams need to establish who can authorize tool access and who can see or share the resulting work before treating an AI participant as a routine addition to calls.

About the Author

John K. Waters is the editor in chief of a number of Converge360.com sites, with a focus on high-end development, AI and future tech. He's been writing about cutting-edge technologies and culture of Silicon Valley for more than two decades, and he's written more than a dozen books. He also co-scripted the documentary film Silicon Valley: A 100 Year Renaissance, which aired on PBS.  He can be reached at [email protected].

Featured