On 28 October Anthropic added three tools to Claude Enterprise so companies can measure what their AI use is delivering. Smart Reports analyse a sample of conversations by team and type of work and show cost, outputs, friction points and recommendations. In-app surveys collect employees' views on the time they save, while pooled budgets give each team a monthly allowance with individual limits as a safeguard. The company stresses that the reports are meant for investment planning, not for evaluating individual employees.

What it means: After two years of enthusiasm, management teams are now asking how much AI costs and what it produces. Anthropic is answering with tools that turn usage into numbers a finance director understands, something its competitors are gradually doing as well. For Greek companies that have handed out AI licences to staff, the main lesson applies whatever the provider: without measurement, subscriptions become one more fixed cost that nobody checks. Pooled budgets per team are a practical fix, because some users consume a lot and others almost nothing. There is also a data protection side. Analysing a sample of conversations, even anonymously and at team level, touches the GDPR and employment law, since it can be seen as a form of monitoring. A company that turns on such features should inform staff in writing, record the purpose of the processing and, where needed, carry out an impact assessment, so that measuring value does not turn into watching individuals.

Read the original on Anthropic

Related service: Custom software

Related stories