Enterprise AI is entering its prove-it era. That was the message from day one of Reuters Momentum AI Austin 2026, as executives moved beyond experimentation to confront a harder question: is AI actually delivering measurable business value?
KPMG's Swami Chandrasekaran, Head of AI and Data Labs, offered one answer: "cost per accepted output." The measure looks beyond the cost of the technology itself to capture what businesses really spend producing usable AI output, including human review, systems for catching errors, and fallback processes when AI fails.
For executives under growing pressure to justify AI budgets, it provides a practical way to distinguish genuine returns from costly experimentation.
Jess Jarvis, Principal at ZS, put the challenge into context: "We're putting AI into our organizations, but we have the same roles, the same organizational structure, the same meetings, the same decision rhythms, the same incentives. Until we redesign that architecture around AI, we're not going to see the full value."
That focus on measurable impact ran throughout the day. Speakers from AGCO and Ancestry urged businesses to be prepared to kill AI projects that fail to demonstrate val...