subscribe
Source linked

Cloudflare CASB surveille maintenant Claude AI comme n'importe quel autre SaaS

blog.cloudflare.com@systems_wire5 days ago·Artificial Intelligence·8 comments

Le CASB de Cloudflare extrait désormais des données de l'API Claude Compliance d'Anthropic, donnant aux équipes de sécurité une visibilité complète dans l'utilisation de l'IA sans agents d'endpoint.

cloudflareanthropicclaudecasbai governanceartificial intelligence

Cloudflare CASB can now read Claude’s internal audit logs and surface DLP violations in the same pane as Microsoft 365.

Why AI Governance Needs a New Model

AI tools differ from SaaS in that they converse, persist, and embed themselves in workflows via APIs. An employee can paste customer data into a prompt, a developer might leave an API key unrotated, and a model can generate content that leaks secrets. Traditional network‑level blocks miss these inside‑app actions, leaving compliance teams blind to the very data that matters.

What the Integration Actually Does

The Claude Compliance API exposes programmatic access to Claude Enterprise and Claude Platform. Cloudflare CASB consumes this endpoint and scans for five categories of findings:

  • Projects – Detect projects shared across the organization or with specific groups.
  • Project attachments – Files added to projects that violate DLP policies.
  • Chat files – User‑uploaded and provider‑generated files that breach DLP.
  • Chat messages – Prompts and responses that contain sensitive data.
  • Artifacts – Provider‑generated documents that violate DLP.

Findings appear in the Cloudflare dashboard, grouped by category and ordered by severity. Read‑only endpoints prevent data loss while still delivering full visibility.

Practical Impact for Enterprise Teams

Security teams can triage, assign, and remediate Claude‑specific risks using the same workflows they use for Microsoft 365, Google Workspace, or Salesforce. No endpoint agents or inline traffic inspection are required; the CASB pulls data directly from Claude’s audit logs. In the near future, support for the Activity Feed will add real‑time alerts. By treating AI as another SaaS app, organizations close the visibility gap that has left them exposed to data leaks and policy violations.

With this API bridge, security teams can finally treat AI as another SaaS app, closing the visibility gap that has left enterprises exposed to data leaks and policy violations.

Source: Announcing Claude Compliance API support with Cloudflare CASB
Domain: blog.cloudflare.com

Read original source ->

External source stays available while the OJO article and comment thread stay local.

More in Artificial Intelligence

view topic
IBM Research Finds Agent Logic Boosts AI Adoption

IBM Research demonstrates that agent logic can improve AI adoption in enterprise workflows by 30×, reducing token consumption and increasing accuracy.

NVIDIA Cosmos 3 Unifies Physical Reasoning and Action in One Model

A single Mixture-of-Transformers architecture replaces separate models for world generation, scene understanding, and robot policy generation.

DuckDuckGo's 30% Install Surge Signals AI-Psychosis in Google's Search

Google's push for AI-augmented search has spurred a 30% jump in DuckDuckGo installs, a clear sign that users are pushing back against the tech giant's new direction.

ESMFold2 Predicts 1.1 Billion Protein Structures

ESMFold2, a new AI tool, has generated an atlas of over 1 billion predicted protein structures, vastly expanding the known protein universe. The atlas eclipses the AlphaFold Database by more than 800 million entries.

Comments load interactively on the live page.