Artificial intelligence company Anthropic has introduced new reporting tools designed to help businesses understand what they are getting in return for spending on its AI products, including its coding assistant, Claude Code.
Announced on Wednesday, the tools use internal analytics and employee surveys to give managers a more detailed view of how teams use AI, what they produce, how much their work costs and where workflows encounter problems.
The reports also provide recommendations to help businesses improve their results. Anthropic says the data will not be used to train its models by default.
The launch comes as companies face growing pressure to justify their AI spending and determine whether the technology is delivering measurable productivity gains.
New tools aim to improve AI returns
Anthropic’s reporting suite allows managers to examine team activity over a selected period, review work outcomes and identify common obstacles in AI-assisted workflows.
The company is also introducing pooled budgets, allowing teams to manage token spending collectively. Tokens are units of text or other data that AI models process when responding to prompts or performing tasks.
The tools could help businesses identify where AI delivers value, where employees encounter difficulties and how workflows can be improved without increasing costs unnecessarily.
For Anthropic, the move also provides an opportunity to strengthen relationships with enterprise customers by helping them optimise their use of its products.
AI coding costs rise as businesses expand adoption
The launch comes amid rapid growth in spending on AI coding assistants.
Jellyfish, a company that provides software engineering management tools, released a report based on activity involving 276,000 users of agentic coding tools. According to the report, Claude Code and other tools powered by Anthropic models accounted for nearly one-third of proposed code changes submitted for review.
Spending on these tools increased tenfold since the beginning of the year, with costs reaching approximately 2% of total engineering expenditure, the report found.
Nicholas Arcolano, Jellyfish’s chief scientist, told TechCrunch that the rapid increase in spending was creating a challenge for chief financial officers who had not anticipated such growth.
He said businesses were reaching a point where simply spending on AI was no longer sufficient; they needed to optimise how the technology was used.
Enterprise customers are harder to win away
Anthropic’s reporting tools also reflect the importance of retaining corporate customers as competition among AI companies intensifies.
According to Nikolas Albarran, Jellyfish’s AI product lead, OpenAI’s Codex gained traction in the company’s data following the release of its Astra model, but the shift had not yet displaced Anthropic’s established position among enterprise users.
Albarran told TechCrunch that Anthropic had benefited from effective enterprise-focused services and established workflows that companies had invested time and resources in developing.
That investment can make switching providers more difficult, even when competing AI models offer similar capabilities.
Businesses must often retrain employees, adapt workflows and account for differences in how AI tools handle files and instructions. These adjustments create additional costs that can discourage companies from moving to another provider.
Reporting tools could strengthen customer loyalty
AI companies are increasingly investing in forward-deployed engineering teams, which place specialists inside customer organisations to help businesses implement and integrate AI systems.
Anthropic’s reporting suite could offer a lighter version of that support by helping companies identify problems and improve results through data-driven recommendations.
The strategy comes as businesses continue to evaluate whether AI adoption is translating into meaningful productivity gains. For Anthropic, demonstrating measurable value could help justify enterprise spending while making its products more deeply embedded in customers’ daily operations.
