Cloudflare — searchable YouTube transcripts & AI Q&A

Introducing Cloudflare OS at Immerse Amsterdam 2026

Cloudflare argues that safe AI adoption requires a platform balancing democratized innovation with central guardrails. They introduce Cloudflare OS, which uses gatekeepers to manage API credentials on demand and allows users to build and share small apps, or gadgets, that keep data within the organization. The platform has already been used internally to build over 4,000 tools, saving thousands of hours and flagging numerous issues before release.

Programmable SASE: The Fastest Path to Safe AI Adoption | Immerse Amsterdam 2026

Cloudflare's SASE solution can be customized using its developer platform to enforce granular AI governance, such as redirecting users from unapproved AI apps to approved ones and blocking sensitive data leaks via DLP policies. The speaker demonstrates how to build a custom worker that requires users to justify access to ChatGPT, and how to secure AI agents by routing MCP traffic through a portal and model calls through an AI gateway, with DLP controls on both paths. The talk concludes by showing how zero-trust policies can integrate custom logic, like requiring employees to complete training before accessing Cloudflare OS.

Navigating the Future of Data Sovereignty in EMEA | Immerse Amsterdam 2026

Digital sovereignty is a global concern, not just a European one, and Cloudflare argues that true sovereignty depends on a multi-vendor landscape and global networks for security. The speaker explains that while there is no final legal answer to sovereignty yet, Cloudflare is building technology for granular data control and engaging with governments to shape policy. He also addresses the 'kill switch' fear, noting that Cloudflare would resist such orders and has a history of taking governments to court.

Secure By Default: Shipping Agents on Cloudflare | Immerse Amsterdam 2026

Cloudflare argues that shipping AI agents securely by default requires treating them like human employees under zero trust principles, with identity, sandboxing, and audit controls. The speaker illustrates this with three personas—Sally, Alex, and Harry—showing how unmanaged AI use can leak data, overexpose APIs, or create dangerous blast radii. Cloudflare's platform provides primitives like Workers isolates, Durable Objects, and Secure Web Gateway to enforce these controls automatically.

Supply Chain Attacks in the SaaS Era: Unpacking the Drift Breach | Cloudflare x The CISO Signal

The Drift breach shows how attackers can inherit trust by compromising a SaaS provider's credentials, turning legitimate integrations into a backdoor into hundreds of organizations' Salesforce environments. The speaker argues that zero trust principles, such as continuous verification and least privilege, are essential to mitigate such risks, and that organizations must move beyond one-time compliance checks to maintain visibility and control over third-party access.