WAF: Updated unsafe topic detection for AI Security for Apps
AI Security for Apps now supports an updated set of categories for detecting unsafe topics in incoming prompts. The cf.llm.prompt.unsafe_topic_categories values have changed, and existing WAF custom rules referencing removed or renamed categories will no longer match.
AI Security for Apps now supports an updated set of categories for detecting unsafe topics in incoming prompts.
The values available in cf.llm.prompt.unsafe_topic_categories have changed. Existing WAF custom rules remain valid, but rules that reference a removed or renamed category will no longer match that category. Review any rules that use this field and update their expressions to use the currently supported values.
For category descriptions and configuration guidance, refer to Unsafe topics.
Source: original entry ↗
More from Cloudflare
Follow Cloudflare to get its new changes in your feed and email digest.
Cloudflare One Client for Windows - Hotfix for split tunnel disconnections
This release resolves an issue where the Windows Cloudflare One Client could repeatedly disconnect and reconnect when a split tunnel configuration with hostname-based entries failed to apply.
Cloudflare One Client for macOS - Split tunnel hostname fix
This hotfix resolves an issue where the Cloudflare One Client for macOS would repeatedly disconnect and reconnect when a split tunnel configuration with hostname-based entries failed to apply.
Cloudflare One Client for Linux - Split tunnel hostname configuration fix
Fixed an issue where the Linux client would repeatedly disconnect and reconnect when a split tunnel configuration with hostname-based entries failed to apply, improving stability for users with complex tunnel setups.