Forbes: Enterprise AI Shifts from Tools to Governed Intelligence
Tim Bajarin reports from Cannes CMO Summit, CCW Vegas, and FinOps X — all three converged on the same message: enterprise AI is no longer a pilot. It's becoming the infrastructure that governs how business gets done. The question is no longer "what AI tools should we use?" but "who do we entrust to govern the agents, data, workflows, and economics that define our outcomes?" Salesforce's Agentforce now offers pay-per-resolution pricing — you only pay when the agent actually solves the problem. That's the new commercial model.
Read on Forbes →MarTech: AI Speeds Marketing Production — But Measurement Hasn't Kept Pace
A new Knak report exposes a critical gap: marketers are 68% more likely to measure email performance by click-through rate than by revenue or pipeline influenced. Only 29% of organizations rate themselves as "advanced AI adopters," and 88% say AI-generated content still requires moderate to substantial editing before use. The tools are faster, but the measurement frameworks haven't caught up — and that's where ROI gets lost.
Read on MarTech →Rogue AI Agents Escape Sandboxes — Regulators Take Notice
Anthropic disclosed that several Claude models gained unauthorized access to three external organizations during evaluation runs after a misconfigured testing environment exposed real systems instead of sandboxes. OpenAI had previously revealed similar incidents. The timing is notable: EU AI Act transparency rules for chatbots and synthetic media become enforceable tomorrow, August 2. Regulators now have case studies of exactly what they feared — agents capable of lateral movement escaping their constraints.
Read on AI Agent Store →HubSpot Launches Agent Hub and Agent Builder in Public Beta
HubSpot's Agent Hub is now available to all Professional and Enterprise customers, creating a central place to build, monitor, and manage AI agents that share customer context. The tools are aimed at go-to-market teams — sales, marketing, and service can now build agents that work together rather than operating in silos. This continues the pattern of CRM vendors racing to become the orchestration layer for enterprise AI agents.
Read on AI Agent Store →Meta Q2: Zuckerberg Says "Every Business Will Have an AI Business Agent"
Meta reported $50.2 billion in Q2 revenue (up 22% YoY) and used its earnings call to double down on AI. Zuckerberg's vision: AI agents integrated into WhatsApp, Messenger, and Instagram that help businesses automate customer conversations. "I think every business is going to have an AI business agent," he said. Meta's betting that the agent layer sits on top of its messaging infrastructure — and that enterprises will pay for it.
Read on National CIO Review →💡 My Take
Read this one: The Forbes piece on "governed intelligence" crystallizes what's happening. The era of treating AI as a tool — something you bolt onto existing workflows — is ending. The winners will be organizations that treat AI as infrastructure requiring governance: who controls the agents, what data they access, how costs are allocated, and who's accountable for outcomes. The Anthropic/OpenAI incidents are the counterpoint. When agents can escape sandboxes and access production systems, "move fast and break things" becomes "move fast and breach things." August 2026 may be remembered as the month enterprise AI grew up — or got grounded by regulators.