Google Cloud Next 2026 Takeaways with James Governor and Kate Holterhoff

This year's #CloudflareConnect is going to be bigger than ever with our fully integrated platform, merging AI-driven security, advanced networking, and developer-led growth.

I don't normally get excited about this type of thing six months in advance, but everything from DNS/AppSec to AI is going through some exciting evolution!

Excited to share part 2 of the PII-shield blog series. Privacy in AI Agents is becoming increasingly critical. In this post, we explore how to implement a Privacy Proxy pattern using Microsoft Agent Framework middleware - helping ensure that sensitive data is protected as it flows through agent-based architectures. This approach introduces a clean way to: - Intercept and sanitize sensitive information - Enforce privacy policies consistently - Build trust into agent workflows by design If you’re working with AI agents or exploring the Microsoft Agent Framework, this pattern can help you balance capability with control. Special callout to my co-author Prafulla Wani for the collaboration and insights on this topic.

Your LLM apps are live in production, but are you actually sure they're behaving the way you expect? RapDev helps teams bring real visibility, guardrails, and cost control to LLM-powered systems using Datadog LLM Observability. From tracing prompts and detecting hallucinations to tracking token costs, we help you turn guesswork into something measurable and controlled. Read the full solution brief: https://rap.dev/4w8JAsr

SecOps stat of the day - To date, our bench of AI agents have performed >25 million investigations 🤯 The leaderboard for product-specific agents: - Microsoft Entra ID 6.3M - Microsoft Sentinel 5.7M - AWS GuardDuty 5M - SentinelOne Endpoint 1.6M - CrowdStrike Falcon Identity 1.2M The image below is an example of an agent's standard operating procedure (SOP), effectively the combined automation and analysis workflow required to reason about a given input.

AI security conversations typically focus on the model, but this upcoming workshop with HashiCorp, an IBM company, and Carahsoft, focuses on the identity layer behind it. The hands-on session will explore how agencies are securing AI-driven environments with dynamic secrets, OAuth 2.0 token exchange and identity-based access controls built around Zero Trust principles, alongside real-world Public Sector use cases spanning multi-cloud and tactical edge environments. Carlos Hernandez from HashiCorp will also share practical guidance on securing modern infrastructure without adding unnecessary complexity. Sign up: https://ow.ly/K8RA30sVMIC

To keep pace with AI workloads, businesses across industries are transforming their data centers. 💡Discover how Cisco is solving new security, infrastructure, and data needs in this fireside chat with Carlos Rojas and Al Slamecka. https://cs.co/6048BBVYEo

𝗕𝗲𝘀𝘁 𝗔𝘂𝘁𝗵𝗲𝗻𝘁𝗶𝗰𝗮𝘁𝗶𝗼𝗻 𝗣𝗹𝗮𝘁𝗳𝗼𝗿𝗺𝘀 𝗳𝗼𝗿 𝗔𝗜 𝗔𝗴𝗲𝗻𝘁𝘀 𝗮𝗻𝗱 𝗠𝗖𝗣 𝗦𝗲𝗿𝘃𝗲𝗿𝘀 𝗶𝗻 𝟮𝟬𝟮𝟲 As MCP crosses 97 million monthly SDK downloads and AI agents move into production workflows, authentication has become the most critical infrastructure decision teams face. This guide ranks the eight leading platforms — WorkOS, Stytch, Auth0 by Okta, Composio, Nango, Arcade, TrueFoundry, and Cloudflare — on spec compliance, enterprise identity depth, integration breadth, and real-world fit for 2026 deployments. The post Best Authentication Platforms for AI Agents and MCP Servers in 2026 appeared first on MarkTechPost. https://lnkd.in/grNjCMiU

I keep thinking about the difference between privacy by policy and privacy by architecture. Privacy by policy means "we promise not to look at your data." Privacy by architecture means "we built it so we physically cannot access your data." GhostLine is the second one. The raw video and audio never leave the guest's device. Our server handles signaling only. We couldn't access the media stream even if someone showed up with a court order.

Connecting the dots across numerous devices, endpoints, and petabytes of data is an ongoing challenge, especially as AI workloads grow. If your primary interest at Cisco Live 2026 is learning how you can use Splunk to build more performant, secure, and stable networks, then this track is for you. https://bit.ly/4fwVciV