Anthropic makes Claude Enterprise smart reports, pooled budgets and in-app surveys generally available for AI ROI tracking
Matéria consolidada. Reúne 2 matérias que o Tifon tinha publicado separadamente sobre a mesma história. A IA reescreveu o relatório com o que as 2 fontes trouxeram em 29/10/2026 00:07; os links antigos trazem para cá.
Fontes: techcrunch.com ↗ zdnet.com ↗
Anthropic moved three Claude Enterprise tracking features, smart reports, pooled budgets and in-app surveys, from beta (released in September) to general availability on 2026-10-28 (zdnet.com). They show enterprise customers what employees use Claude for, what it costs and how spend is distributed. TechCrunch (techcrunch.com) describes the same launch as a suite of reporting tools for enterprise customers of the coding tool Claude Code, showing what teams produced, what it cost and where workflows broke down. ZDNet links the release to Anthropic's push to sustain enterprise growth after it filed its IPO on Monday (zdnet.com). This is product and business news: neither source describes a vulnerability, threat actor or security incident. For defenders, the relevance is that these features add usage and spend telemetry about employee AI use.
Detalhes técnicos
- Smart reports (off by default) track which tasks employees run with Claude, how often and at what cost. They also record where sessions hit friction and where routine workflows could be created. Anthropic's documentation says they should not be used to evaluate individual performance or make employment decisions (zdnet.com).
- Pooled budgets set a monthly spend cap for a group, on top of each seat's individual monthly limit. Admins can base limits on prior monthly spend. Users can belong to several pooled budgets, ordered by priority (zdnet.com). TechCrunch also reports pooled group budgets for token spending (techcrunch.com).
- In-app surveys are the third feature; ZDNet gives no details on them (zdnet.com). TechCrunch mentions employee surveys and internal analytics that let managers review a team's output over a date range and get improvement recommendations (techcrunch.com).
- Anthropic says it will not use this data for model training by default (techcrunch.com).
- Pricing cited: Claude Enterprise is $20 per seat per month plus API token rates. Claude Teams is $20 per seat, capped at 150 users, with tokens included (zdnet.com).
- Market context: Ramp data cited by ZDNet shows Anthropic spend peaked in July and is slightly lower so far in October (zdnet.com). TechCrunch reports an operating profit in Q2 and 900% net revenue retention for Anthropic. It also cites a Jellyfish report covering 276,000 customers that found Anthropic-based tools behind nearly a third of proposed code changes, with spend up 10x since the start of the year and equal to about 2% of engineering cost. Jellyfish says OpenAI's Astra model and Codex have not slowed Anthropic's adoption (techcrunch.com).
Relatos divergentes
- Product scope: ZDNet presents the release as general availability of Claude Enterprise features (zdnet.com). TechCrunch presents it as reporting tools for Claude Code enterprise customers (techcrunch.com). Both describe the same launch, and the sources do not say whether the features cover both products.
- Adoption trend: ZDNet cites Ramp data showing Anthropic spend slightly down in October after a July peak (zdnet.com). TechCrunch cites Jellyfish data showing spend up 10x since the start of the year (techcrunch.com). The two use different datasets and periods, so they are not directly comparable.
Linha do tempo
-
Anthropic launches Claude Code reporting tools and pooled token budgets to show enterprises AI spend ROI
ler a análise desta fonte ocultar
Anthropic announced a suite of reporting tools for enterprise customers of its AI coding tool Claude Code. The tools show what teams produced, what it cost and where workflows broke down. This is a business and product announcement with no security incident, vulnerability or threat activity described. Managers can use internal analytics and employee surveys to review a team's output over a date range and get improvement recommendations. Anthropic says it will not use this data for model training by default, and it is adding pooled group budgets for token spending. TechCrunch reports that Anthropic recorded an operating profit in Q2 and 900% net revenue retention. A Jellyfish report covering 276,000 customers found Anthropic-based tools behind nearly a third of proposed code changes, with spend up 10x since the start of the year and equal to about 2% of engineering cost. Jellyfish also says OpenAI's Astra model and Codex have not slowed Anthropic's adoption.
-
Anthropic makes Claude Enterprise smart reports, pooled budgets and in-app surveys generally available for AI ROI tracking
ler a análise desta fonte ocultar
Anthropic has moved three Claude Enterprise tracking features from beta (released in September) to general availability: smart reports, pooled budgets and in-app surveys. They give enterprise customers more visibility into what employees use Claude for, what it costs and how spend is distributed. The article frames the release as an effort to sustain enterprise growth, since Anthropic filed its IPO on Monday. This is product and business news with no vulnerability, threat actor or incident involved.
Technical Details
- Smart reports (off by default) track which tasks employees run with Claude, how often and at what cost. They also record where sessions hit friction and where routine workflows could be created. Anthropic's documentation says they should not be used to evaluate individual performance or make employment decisions.
- Pooled budgets set a monthly spend cap for a group, on top of each seat's individual monthly limit. Admins can base limits on prior monthly spend. Users can belong to several pooled budgets, ordered by priority.
- In-app surveys are the third feature, but the article gives no details on them.
- Pricing cited: Claude Enterprise is $20 per seat per month plus API token rates. Claude Teams is $20 per seat, capped at 150 users, with tokens included.
- Ramp data cited by the article shows Anthropic spend peaked in July and is slightly lower so far in October.