Anthropic will give managers reports on employees’ work with Claude

Anthropic has introduced Smart Reports, a tool that lets managers analyze how employees use Claude and the company’s other products. The system generates summaries of workplace interactions with the chatbot and shows costs as well as the results of its responses. Anthropic says sessions are anonymized, but it is not yet clear how reliable that protection is in practice. The new feature could help the company attract enterprise customers: according to available figures, they account for 80% of its business, with individual subscriptions making up the rest.

Anthropic will give managers reports on employees’ work with Claude

What Smart Reports shows

According to Anthropic’s support page, managers will be able to generate reports on how their team uses Claude and track the work completed each day. The tool covers more than individual chats: it can sort employees’ AI output by overall cost and by chatbot responses. This gives management a way to assess how effectively employees use AI tools.

Anthropic says the system anonymizes sessions. However, it is not yet known how reliably it protects data in practice. In the past, another Anthropic project involving data anonymization led to a privacy problem: the information was re-identified using widely available large language models.

Why it matters to the company

Smart Reports arrives as Anthropic seeks to sign more enterprise contracts ahead of a potential stock market listing. According to available figures, enterprise customers account for 80% of the company’s business, with individual subscriptions making up the rest.

Anthropic’s success in the public markets will depend to a significant extent on how effectively the company can attract enterprise customers with products such as Smart Reports. The new tool’s practical value has yet to be established.

Dataist’s take: Smart Reports help managers monitor the costs and results of Claude use, but also expand their ability to oversee employees’ work. It remains unclear how convincing the promises of data anonymization will be; the source recalls problems with Anthropic’s previous attempt to protect privacy in this way.