Salesforce: AI Agent Adoption Tripled in 2026
The new Agentic Enterprise Index shows the average number of AI agents per organization grew from 5 to 13 since early 2025 — a 160% increase. Agent creation time dropped 53% to just 1.9 days. Employee-initiated agent sessions surged 3x as trust deepens. Retail leads with 9 unique actions per agent.
Read on ZDNet →Deloitte: Only 5% of Enterprises Ready for AI Agents
New Deloitte survey of 500+ leaders reveals a massive expectations-vs-readiness gap. 74% expect half their processes redesigned around agents within four years — but only 5% say they're prepared today. Just 15% have scaled multi-agent deployment, and half aren't investing in workforce transformation.
Read on Deloitte →OpenAI: Frontier Firms Generate 8.3x More Output
New OpenAI research shows enterprise AI is shifting "from assistance to execution." Frontier firms (top 10% by usage) now generate 8.3x as many output tokens per user as typical firms — up from 2.6x in January. Codex accounts for 64% of enterprise output. Agents are spreading fastest in legal (108x), sales (41x), and marketing (26x).
Read on OpenAI →MIT Tech Review: Data Leaders Trust Their Agents 100%
A new study finds organizations providing 70%+ data access to agents ("data leaders") have 100% trust in agent decisions — vs. ~50% average. Two-thirds of data laggards say legacy systems limit AI scaling. The message: trustworthy data is the foundation for agentic AI. Without it, agents can't act at speed.
Read on MIT Tech Review →Forbes AI 50: Anthropic Hits $380B, Gamma Reaches 100M Users
The 2026 Forbes AI 50 highlights: Anthropic now valued at $380B after Claude surpassed ChatGPT downloads. Gamma (AI presentations) hit $2.1B with 100M users — from "worst idea ever" to profitable in 3 years. Reflection ($8B) is the U.S. answer to DeepSeek, backed by Nvidia and 1789 Capital.
Read on Forbes →💡 My Take
Read this one: The Deloitte survey. The 5% readiness figure is sobering — but the real story is that 74% of leaders expect half their processes to be redesigned around agents within 4 years. That's not experimental anymore. The gap between expectation and readiness is where PMM leaders can differentiate.