Check out our new GigaOm Radar! Learn more

Cloudflare Connect 2026 Retrospective

Eric Phenix Cloudflare Connect Selfie


I recently attended Cloudflare Connect in San Francisco— an incredible conference that fundamentally reshaped my understanding of Cloudflare’s role in the cloud ecosystem.

I’ve been deep into cloud infrastructure since 2016. In those 10 years, I’ve worked in Microsoft, Google, Amazon, and Oracle clouds. Before GigaOm, I worked for AWS from 2019 to 2023. Cloudflare has always been the thing you put in front of your application running on these clouds to make it more secure and resilient. This is no longer the case.

The core message at this year’s event was clear: AI is actively defining the future of software development and hosting. Cloudflare isn’t building for yesterday’s infrastructure patterns—it isn’t just a target for lift-and-shift legacy migrations, but the platform where the next generation of applications will be built.

Agentic tooling and AI threat defense took center stage. Cloudflare sees over 300 billion attacks daily. From the publication of an exploit to its first detected malicious use, Cloudflare estimates organizations have just 22 minutes. Their question to the audience—“Can you patch your systems in 22 minutes?”—underscored the need for continuous remediation and agentic defense. As AI tools like Mythos uncover vulnerabilities faster, maintaining security now requires continuous, automated agentic red-teaming.

The conversation was not limited to security. Our productivity model is fundamentally changing, and the shift is again catalyzed by AI. At Cloudflare, the majority of code is now being shipped by directors, VPs, and above. To handle the AI sprawl and widespread AI use, they built a harness for AI governance and centralization called Cloudflare OS. It provides necessary guardrails and access controls around AI tooling, preventing unmonitored code execution and sandbox breakouts. This tool was recently open-sourced and is available for everyone to deploy for free. 

During our Analyst Roundtable discussions, Cloudflare leadership highlighted several key security metrics and architectural shifts:

  • Agentic Defense & Rapid WAF Generation: As AI tools accelerate vulnerability discovery, Cloudflare built automated pipelines that process code vulnerabilities and automatically deploy targeted Web Application Firewall (WAF) rules within six weeks.

  • Custom Vulnerability Database: Recognizing that legacy databases couldn't keep pace with frontier models, Cloudflare engineered a custom vulnerability database in just three days, directly supplying context to coding agents for automated remediation.

  • Massive Scale & Threat Defense: Cloudflare's global network processes over 300 billion attacks daily and recently mitigated a record 31.4 Terabits per second (Tbps) DDoS attack.

  • Internal AI Ecosystem: Cloudflare’s CISO manages over 450 AI assets, skills, and agents, driving an 8x throughput increase across GRC, EDR, and security operations.

Here are key resources and blog posts detailing Cloudflare's latest announcements:

Discover what GigaOm can do for you

Independent, practitioner-led research that helps you choose, prove, and sell enterprise technology with confidence.

[Local dev] blog_closing_band_hubspot_form_id is not configured — the embedded HubSpot form would appear here.