Introducing Cloudflare Application Confidence Score For AI Applications

Cloudflare announces Application Confidence Scores—part of its AI Security Posture Management in the Cloudflare One SASE platform—to help Security, IT, Legal, and GRC teams assess SaaS and Gen-AI applications using a transparent, objective rubric based on publicly available evidence. The system computes two 5-point scores per account tier—Application Confidence (compliance, data retention and third-party sharing, security controls and transparency, financial stability) and Gen-AI Confidence (ISO 42001, deployment security, model card presence, and whether user prompts are used for training)—via automated ingestion and human review. Example scores for providers like OpenAI, Google Gemini, and Anthropic illustrate tier-based differences, and Cloudflare plans to publish these in an Application Library and integrate enforcement into Gateway and Access with DLP/RBI controls, inviting community feedback to refine the rubric.