Every CIO will have to answer for every token | Factory's Matan Grinberg
Enterprise CIOs will soon be accountable for every AI token spent, role by role
“Every CIO is going to need to answer for every incremental token, where do we put it?”
Factory's Matan Grinberg argues that blanket token caps across an enterprise are a near-term problem CIOs haven't solved yet but will have to within 12 months. Factory's answer is a dynamic model router that allocates different models—Gemini Flash for low-stakes vibe-coding, fine-tuned models for COBOL, Anthropic for testing—based on task and org role. The signal here is the emergence of token governance as a CFO/CIO-level infrastructure concern, not just a developer cost optimization.