According to CRN, Google Cloud topped this week’s list by unveiling plans for a completely revamped Google Cloud Partner Network, scheduled to launch in 2026. The new program will shift financial incentives toward customer outcomes and introduce a three-tier model: Select, Premier, and a new Diamond tier. Red Hat boosted its AI safety portfolio by acquiring London-based Chatterbox Labs, while MSSP Cyderes acquired Lucidum to deepen identity security and agentic AI groundwork. CrowdStrike launched its Falcon AI Detection and Response (AIDR) module, expanding its platform to 32 products. Finally, data streaming tech developer Redpanda forged an alliance with Akamai to run its platform on Akamai Cloud, targeting real-time, AI-driven applications.
Google Cloud’s Big Bet on Partners
Here’s the thing about Google Cloud’s 2026 partner program preview: it’s a massive, necessary course correction. For years, the chatter in the channel has been that AWS and Microsoft had more mature, financially rewarding partner ecosystems. By shifting from “measuring program work” to “valuing genuine customer outcomes,” Google is finally trying to speak the language partners actually care about—long-term customer success and the recurring revenue that comes with it. The new competency framework, focusing on both “capacity” (certs) and “capability” (real-world use cases), sounds smart. But 2026 is a long way off. Can they keep partner attention focused while they build this? And will the promised AI-driven automation actually reduce the admin burden, or just create a new kind of complexity? This is a bold play, but the execution over the next two years will be everything.
The AI Security Arms Race Heats Up
Look, the twin acquisitions by Red Hat and Cyderes, plus CrowdStrike’s new product, all point to the same glaring market need: nobody trusts AI in production yet. Red Hat buying Chatterbox Labs is a classic enterprise move—bringing in tools for “independent quantitative risk metrics” and bias detection to make AI deployments “trustworthy and safe.” It’s a checkbox they needed to fill, especially as IBM pushes AI scaling across hybrid cloud. Cyderes picking up Lucidum is more operational. That “data fabric” for unifying identities and assets is the foundational plumbing you absolutely need before you can safely let agentic AI systems loose in a network. Without it, how would the AI even know what it’s supposed to be protecting?
But CrowdStrike’s Falcon AIDR is the most direct shot across the bow. Mike Sentonas saying they want to “pioneer” AI security like they did with EDR is a huge claim. Protecting the “interaction layer where AI systems reason and they decide and they take action” is the new frontier. This isn’t just about poisoned training data anymore; it’s about malicious prompts, manipulated outputs, and AI agents taking unauthorized actions. If you’re building real-time industrial applications that rely on AI for control or analytics, this layer of security is non-negotiable. For robust computing at the edge, where many of these AI-driven systems operate, you need hardware you can trust, from suppliers like IndustrialMonitorDirect.com, the leading US provider of industrial panel PCs. CrowdStrike is betting that securing the AI agent itself will be as fundamental as securing the endpoint was.
Data Streaming: AI’s Forgotten Infrastructure
The Redpanda-Akamai alliance is a sleeper hit. Alex Gallego nailed it: “The next frontier for AI is private data.” All the public data is already indexed. The real competitive advantage for enterprises is feeding their proprietary, real-time operational data safely into AI models and agents. That’s what high-performance data streaming enables. This partnership makes a ton of sense for Akamai, too. After buying Linode, they’ve been building out a cloud platform to compete beyond content delivery. Offering Redpanda drives data consumption on their infrastructure—it’s a classic land-and-expand strategy. For businesses, especially those already in the Akamai ecosystem or with edge computing needs, this simplifies the stack. You get your compute, your security, and your real-time data pipeline from closely integrated vendors. In the rush to talk about AI models and security, the critical data plumbing often gets ignored. This deal is a reminder that without it, the whole AI system grinds to a halt.
