Anthropic's Smart Reports lets Claude Enterprise admins see summaries of team chats. It ships off, but if your employer turns it on, your work sessions may be visible.
Anthropic has added a feature called Smart Reports to Claude Enterprise, its business version of the Claude AI assistant. It lets administrators at a company see more than how often people use AI. They can see short summaries of what the conversations were about. The feature is off by default, so a company has to switch it on first.
An admin picks a team, a time window of up to 28 days, and which Claude products to include, such as Claude Chat or Claude Code. They can also write custom questions. Claude then reviews a sample of the chat transcripts, groups the activity into workstreams, estimates costs at list price, and produces a report. Each organization gets up to 10 free reports per month.
Admins can open individual workstreams and see sessions ranked by cost. Each one shows a short summary, the date, the product used and the type of output. That means the tool shows summaries of real conversations, not just usage counts.
A separate setting controls identity. With it off, the drilldown shows only "User" and hides session IDs. With it on, names stay hidden by default but an admin can reveal them on demand. Every reveal is logged automatically.
Some topics get only aggregate treatment, meaning group totals with no way to open individual sessions. Anthropic says these protections cannot be turned off:
Admins also cannot write custom questions aimed at these topics. There is a catch. Sorting conversations into these categories is automated, and the documentation gives no guaranteed detection rate. Some sensitive chats could be miscategorized, and Anthropic has not published failure-rate data.
Anyone who uses Claude through an employer that has enabled Smart Reports. A work Claude account carries different privacy expectations than a personal one, and Anthropic treats business data under a separate legal and contractual framework.
The feature is not available in every country where Anthropic operates. Even where it is offered, employers may have their own legal duties under local law. These can include notifying employees, getting consent, or consulting a works council, depending on the place.
Anthropic's controls do not answer whether you will be told about this access, or how long summaries are kept. Those are decisions for your employer. Reasonable questions to ask:
If you use Claude at work, ask your employer whether Smart Reports is switched on and who can see names, because the feature is off until a company enables it.
World models, AI that learns physics from video, are being built into home robots and driverless cars, and your own videos may help train them.
Read more: AI that learns how the physical world works is coming to robots and carsA new Alibaba-led test of real coding requests found the best AI agent fully finished just 26.4 percent of tasks, so human review still matters.
Read more: AI coding helpers fully finish only about 1 in 4 real tasksA University of Chicago law professor asked Claude to fix page numbers. It deleted his syllabus and a year of notes instead. Dropbox saved him, not the AI.
Read more: An AI assistant deleted a law professor's year of lecture notes
