Anthropic is giving its Claude AI model new skills in data governance and analytics, aiming to let businesses build self-service tools without a data science team. The update lands as the company weighs a potential initial public offering, sources familiar with the matter say.
What Claude’s new skills do
Claude can now enforce data access rules, tag sensitive fields, and generate dashboards from natural language queries. The model handles permissions checks before running any analysis, so a sales manager can see team revenue but not individual salaries. Anthropic says this removes a bottleneck: companies often have the data but lack the budget to hire dedicated analysts.
The governance layer sits on top of existing databases. That means a business doesn’t have to migrate its data to a new warehouse. Claude reads the schema, applies the policies the admin sets, and lets any employee ask questions like “show me last month’s churn by region.”
Enterprise deals are critical for Anthropic’s revenue story. The company has raised billions from investors including Google and Amazon, but it needs to show recurring software revenue before going public. Self-service analytics is a proven wedge into corporate budgets — Salesforce’s Tableau and Microsoft’s Power BI each generate billions annually.
Claude competes with OpenAI’s ChatGPT Enterprise and Google’s Gemini. Neither has pushed hard into governed self-service analytics yet, though both offer data analysis features. Anthropic is betting that “safe” AI — one that respects access controls by default — will win over regulated industries like healthcare and finance.
The IPO timing isn’t set. The company has met with bankers informally, but no filing has been made public. The analytics push gives would-be investors a product story beyond the raw model improvements.
Claude’s new capabilities are rolling out now to enterprise customers on the Pro and Team plans. Anthropic hasn’t said when they’ll reach the free tier.
