Articles from Source: Cloudflare-Blog

State-of-the-art image generation Leonardo models and text-to-speech Deepgram models now available in Workers AI

2025-08-27 14:00
🚀 Exciting updates in Workers AI! We've partnered with Leonardo.Ai and Deepgram to enhance our model catalog. Users can now access advanced image generation and real-time text-to-speech models, all hosted on Cloudflare's infrastructure. This integration allows developers to create applications with low-latency image generation and audio voice agents effectively. Discover more about our new models and how they can support your AI projects! #AI #WorkersAI #Leonardo #Deepgram #Cloudflare
Source: Cloudflare Blog
Nikhil Kothari

Securing the AI Revolution: Introducing Cloudflare MCP Server Portals

2025-08-26 14:05
🚀 Exciting news in AI security! Cloudflare has launched MCP Server Portals in Open Beta. This new feature allows organizations to centralize, secure, and monitor every MCP connection. The Model Context Protocol (MCP) enables Large Language Models (LLMs) like ChatGPT to interact with various applications, enhancing their functionality. Discover how MCP transforms AI into collaborative teammates while addressing security concerns. 🔒 #AI #Cloudflare #Cybersecurity #MCP #TechInnovation
Source: Cloudflare Blog
Kenny Johnson

Best Practices for Securing Generative AI with SASE

2025-08-26 14:00
🚀 As businesses adopt Generative AI, Security and IT leaders face new challenges. Executives push for quick implementation, while teams must create effective AI Security Strategies. 🔒 Cloudflare’s SASE architecture offers a solution, integrating networking and security for secure access to corporate resources. Their platform supports safe AI usage without hindering innovation. 🛡️ Recent enhancements include shadow AI reporting, confidence scoring for AI providers, and protections against...
Source: Cloudflare Blog
Yumna Moazzam

Block unsafe prompts targeting your LLM endpoints with Firewall for AI

2025-08-26 14:00
🚀 Cloudflare has enhanced its AI security suite with the introduction of unsafe content moderation through Firewall for AI. This new feature aims to protect AI applications from malicious prompts that could compromise sensitive data or user trust. 🔒 With this integration, customers can now detect and block harmful prompts at the network level without modifying their application code. Firewall for AI supports various models, ensuring consistent protection across all platforms. 🛡️ This update...
Source: Cloudflare Blog
Mathias Deschamps

ChatGPT, Claude, & Gemini security scanning with Cloudflare CASB

2025-08-26 14:00
🚀 Cloudflare CASB now offers security scanning for generative AI tools like ChatGPT, Claude, and Gemini. Organizations can connect their accounts to detect misconfigurations, data exposure, and compliance risks quickly and without additional software. This integration enhances visibility and helps IT teams manage the security challenges of AI adoption effectively. 🔍 #Cloudflare #CyberSecurity #GenerativeAI #DataProtection #Compliance
Source: Cloudflare Blog
Alex Dunbrack

Introducing Cloudflare Application Confidence Score For AI Applications

2025-08-26 14:00
🌐 Cloudflare has introduced Application Confidence Scores for Gen AI applications to help businesses assess risks associated with shadow IT. These scores will aid security, IT, and legal teams in evaluating third-party applications without needing extensive individual audits. This initiative aims to balance productivity with security, promoting smarter controls rather than blanket bans. Stay informed about the new features in Cloudflare's AI Security Posture Management within the Cloudflare...
Source: Cloudflare Blog
Sharon Goldberg

Unmasking the Unseen: Your Guide to Taming Shadow AI with Cloudflare One

2025-08-25 14:05
🚨 Concerned about data leaks from unsanctioned AI use? The rise of "Shadow AI" poses a significant risk as employees unknowingly expose sensitive information through AI tools. Traditional security measures often miss these interactions. Cloudflare One offers a solution with its Shadow IT Report, providing visibility into AI application usage and helping organizations manage potential risks effectively. Stay informed and secure! 🔍💻 #CyberSecurity #ShadowAI #Cloudflare #DataProtection #Innovation
Source: Cloudflare Blog
Joey Steinberger

Beyond the ban: A better way to secure generative AI applications

2025-08-25 14:00
Generative AI is transforming workplace productivity, but it also poses significant data security risks. 🛡️ Many employees inadvertently share sensitive information with AI tools. Recent incidents highlight the consequences of this, including private chats being indexed online. To address these risks, Cloudflare introduces AI prompt protection, part of its Data Loss Prevention (DLP) system. This feature helps organizations manage AI use while securing sensitive data. AI prompt protection...
Source: Cloudflare Blog
Matt Davis

Bringing Cloudflare’s AI to FedRAMP High

2025-08-25 14:00
🚀 Cloudflare is set to enhance AI capabilities for the US public sector by bringing its AI Developer suite, including Workers AI, AI Gateway, and Vectorize, into FedRAMP Moderate and High boundaries by 2026. This initiative aims to streamline generative AI use while ensuring compliance and security. Workers AI will provide serverless inference close to users, improving response times for agency services. With these tools, agencies can modernize their operations and improve efficiency in...
Source: Cloudflare Blog
Wesley Evans

Cloudflare Launching AI Miniseries for Developers (and Everyone Else They Know)

2025-08-25 14:00
🚀 Cloudflare has launched "AI Avenue," a miniseries designed to demystify AI for developers and the general public. Each episode explores AI's growth, capabilities, and human augmentation stories, making the technology approachable and engaging. Viewers can interact with AI hands-on, breaking down fears and showcasing its potential. Join the journey to learn more at aiavenue.show! #AI #Cloudflare #Innovation #Technology #HumanAugmentation
Source: Cloudflare Blog
Craig Dennis

Welcome to AI Week 2025

2025-08-24 16:00
🌐 Welcome to AI Week 2025! AI is transforming industries by enhancing productivity. Customer support agents can now handle up to ten times more tickets, while software engineers focus on reviewing AI-generated code. This shift allows sales teams to prioritize relationship building over administrative tasks. However, the rapid advancement of AI presents governance challenges. Security concerns arise as sensitive data can be mishandled or exposed. Companies face tough decisions on managing AI...
Source: Cloudflare Blog
James Allworth

MoQ: Refactoring the Internet's real-time media stack

2025-08-22 14:00
🌐 Media over QUIC (MoQ) is a new IETF standard transforming real-time media streaming by unifying various protocols. For years, we relied on RTMP, HLS, and WebRTC, each addressing specific needs but creating complexity in the process. MoQ aims to streamline this by combining low latency with scalability and operational simplicity. Cloudflare has launched the first MoQ relay network, collaborating with industry leaders like Meta and Google to create a shared foundation for future applications....
Source: Cloudflare Blog
Renan Dincer

Cloudflare incident on August 21, 2025

2025-08-22 00:58
On August 21, 2025, Cloudflare experienced significant congestion due to a surge of traffic directed at its clients hosted in AWS us-east-1. 🌐 The congestion began at 16:27 UTC and peaked until 19:38 UTC, affecting user connections with high latency and packet loss. This issue was localized and not a result of an attack. The traffic overload was traced back to a single customer. ⚠️ Cloudflare is working on measures to prevent similar incidents in the future. For more details, check out the...
Source: Cloudflare Blog
Bryton Herdes

How a volunteer-run wildfire site in Portugal stayed online during DDoS attacks

2025-08-21 17:28
🌍🔥 In 2015, fogos.pt began as a volunteer side project to improve access to wildfire data in Portugal. Today, it serves as a vital resource for citizens, media, and government during fire seasons. During the intense 2025 wildfire season, fogos.pt faced DDoS attacks but was protected by Cloudflare's Project Galileo. This platform provides real-time data and is crucial for public safety. Volunteers and partners keep this initiative running without formal funding. #WildfireTracker...
Source: Cloudflare Blog
João Tomé

How we built AI face cropping for Images

2025-08-20 14:00
🚀 Exciting news from Developer Week 2024! We launched AI face cropping, a feature that automatically crops images around detected faces. This marks a significant step in our AI image manipulation capabilities. Our transition from a CPU-based prototype to a GPU-based implementation allows for better performance and scalability, addressing previous technical challenges. This feature is aimed at enhancing image handling for social media platforms and e-commerce sites, making it easier to create...
Source: Cloudflare Blog
Diretnan Domnan

Announcing the Cloudflare Browser Developer Program

2025-08-18 14:00
🌐 Exciting news from Cloudflare! They have launched the Browser Developer Program, aimed at enhancing collaboration between Cloudflare and browser development teams. This initiative focuses on improving security measures while ensuring a seamless user experience. Browser developers can join to share feedback, access best practices, and engage in discussions with Cloudflare engineers. The program also offers insights into updates that may affect browser interactions with Cloudflare Challenges...
Source: Cloudflare Blog
Oliver Payne

MadeYouReset: An HTTP/2 vulnerability thwarted by Rapid Reset mitigations

2025-08-14 22:03
🚨 A new HTTP/2 vulnerability named MadeYouReset (CVE-2025-8671) has been disclosed by researchers at Tel Aviv University. This denial-of-service (DoS) vulnerability affects unpatched HTTP/2 servers, potentially leading to resource exhaustion. If you use Cloudflare for DDoS mitigation, your systems are already protected. MadeYouReset exploits server-sent stream resets, similar to the previously identified Rapid Reset vulnerability. Implementations are encouraged to monitor and limit the use of...
Source: Cloudflare Blog
Evan Rittenhouse

Aligning our prices and packaging with the problems we help customers solve

2025-08-11 23:03
🌐 Cloudflare is introducing Externa and Interna, two new packages designed to simplify how you connect and protect your infrastructure. These use-case bundles address the need for straightforward solutions and pricing, responding to customer feedback for simplicity. Externa focuses on connecting and protecting public-facing systems, while Interna secures internal resources. These packages aim to reduce complexity and enhance security. #Cloudflare #Cybersecurity #TechSolutions #Simplicity...
Source: Cloudflare Blog
Phil Winslow

Redesigning Workers KV for increased availability and faster performance

2025-08-08 13:00
Cloudflare has redesigned its Workers KV service following a significant outage on June 12, 2025. This incident was caused by a failure in the underlying storage infrastructure. To enhance availability and performance, Cloudflare has transitioned to its own storage infrastructure. This change eliminates single points of failure and reduces reliance on third-party providers for redundancy. The updated design aims to provide high availability while improving response times for critical...
Source: Cloudflare Blog
Tyson Trautmann

Partnering with OpenAI to bring their new open models onto Cloudflare Workers AI

2025-08-05 21:05
🚀 Exciting news! OpenAI's latest open-source models are now live on Cloudflare Workers AI. Developers can access two new models: a 120 billion parameter and a 20 billion parameter version. These models feature a Mixture-of-Experts architecture for efficient performance and are designed for transparency and flexibility. 🛠️ In addition, Cloudflare supports the new Responses API, enhancing interaction capabilities. The Code Interpreter and Web Search features are also on the way! #OpenAI...
Source: Cloudflare Blog
Ashish Datta

Reducing double spend latency from 40 ms to < 1 ms on privacy proxy

2025-08-05 13:00
We recently improved the performance of our privacy proxy service by reducing double-spend check latency from 40 ms to less than 1 ms. 🚀 This enhancement helps users browse the web securely without compromising their privacy. It also boosts the efficiency of our service, as we handle millions of requests each second. 🔒 Using a tracing platform and metrics, we identified bottlenecks and optimized our processes. This change is part of Cloudflare's commitment to making the Internet faster for...
Source: Cloudflare Blog
Ben Yang

Perplexity is using stealth, undeclared crawlers to evade website no-crawl directives

2025-08-04 13:00
Perplexity, an AI-powered answer engine, is reportedly using stealth tactics to bypass website no-crawl directives. They modify user agents and change IPs to continue crawling despite explicit blocks. This behavior contradicts the trust-based principles of the internet, which emphasize transparency and adherence to website preferences. As a result, Perplexity has been de-listed as a verified bot and is facing new blocking measures. #Crawling #AI #WebSecurity #DataPrivacy #TechNews 🕵️‍♂️🔍🚫
Source: Cloudflare Blog
Reid Tatoris

Vulnerability disclosure on SSL for SaaS v1 (Managed CNAME)

2025-08-01 13:00
Cloudflare has announced a vulnerability disclosure for its SSL for SaaS v1 (Managed CNAME) product, identified through its bug bounty program. Despite deprecating this version in 2021, some customers have yet to migrate to SSL for SaaS v2. Cloudflare is working to assist these remaining clients in the transition. The architecture of SSL for SaaS v1 posed security risks due to its reliance on IP-based routing without domain ownership verification, leading to potential vulnerabilities. 🌐🔒...
Source: Cloudflare Blog
Ross Jacobs