Salesforce Winter '27: AI Agents Now Run Entire Workflows
Salesforce's Winter '27 release marks a fundamental shift: agents now run complete workflows — qualifying pipeline, resolving service cases, booking appointments — from start to finish. The "Agentic Segmentation and Activation" feature lets marketers describe target audiences in plain language while agents build and verify audiences, then orchestrate campaigns across ad platforms. Early results show Autonomous Scheduling cuts 15-minute booking calls to seconds, and Agentic Commerce Search drives 13% conversion lift.
Read on Salesforce →Yale: The Next AI Battleground Is Data, Not Models
A compelling CNBC op-ed from Yale's Chief Executive Leadership Institute argues the "SaaSpocalypse" narrative is nonsense. The key insight: when intelligence becomes commoditized, value migrates to its scarce complements — trusted proprietary data. Salesforce's blowout earnings (margins at 34.1%, Data 360 ingesting 104 trillion records, up 355% YoY) prove the thesis. Nine of the top 10 AI companies now run on Salesforce — including the frontier labs themselves.
Read on CNBC →Anthropic + OpenAI Combined ARR Tops $115 Billion
The two frontier AI leaders have reached staggering scale. Anthropic's run rate hit $65B by July 2026 (from $9B at year-end 2025), posting its first-ever operating profit in Q2. The combined market concentration is raising eyebrows: ~70% of AI-related revenue at Microsoft, Amazon, Google, and Oracle comes from just these two companies. Both are eyeing IPOs as enterprise adoption accelerates.
Read on Crypto Briefing →Madrona's IA40: AI Shifts From Experimentation to Enterprise ROI
The 2026 Intelligent Applications 40 list highlights a market shift from AI experimentation to enterprise-proven ROI. The 58% repeat winner rate (up from 33% last year) signals growing durability. Databricks remains the only company to appear every year since 2021. The pattern: agentic applications for single professional workflows are maturing into infrastructure layers that make them possible at scale.
Read on Financial Post →MarTech Warning: AI Agent API Costs Are Breaking Budgets
Autonomous marketing agents promise massive labor savings, but a new MarTech analysis warns that ballooning API fees and middleware costs can easily break technology budgets. The hidden loophole: as agents scale, so do the per-call costs to frontier models — and without governance, a single runaway workflow can burn through monthly allocations in hours. The recommendation: build cost caps and usage monitoring before deploying at scale.
Read on MarTech →💡 My Take
Read this one: The Yale/CNBC piece on data as the moat. It's the clearest articulation of where value is shifting. Frontier model capabilities are converging and commoditizing — but the companies that own trusted, proprietary data are the ones AI agents literally cannot function without. Salesforce betting $25B on this thesis with its buyback isn't speculation; it's recognition of structural advantage.