IBM Think 2026

IBM just named something most enterprises are still struggling to articulate.

At Think 2026, CEO Arvind Krishna didn't announce a product. He announced a posture — the AI Operating Model — and declared that the companies figuring it out right now will define the next decade of business.

The distinction matters. For three years, "AI strategy" meant running pilots. Picking tools. Forming task forces. It was innovation theater, and most organizations knew it.

What IBM is describing is different. It's AI embedded into the operating model itself — not layered on top of workflows, but restructuring how work gets done at the foundation. New products like IBM Bob (an AI-native dev agent) and watsonx Orchestrate for multi-agent coordination aren't features. They're infrastructure bets on the idea that the enterprise OS is being rewritten.

JPMorgan's recent disclosure adds weight to this framing. They moved AI from discretionary R&D to core infrastructure spend — and reported $2B in operational savings that have already self-funded the investment. That's not a pilot result. That's a balance sheet decision.

The companies winning in this next phase won't be the ones with the most AI tools. They'll be the ones that redesigned their operating model around AI before their competitors did.

IBM is telling enterprises: the architecture decisions you make in 2026 are the ones you'll be scaling in 2028.

Are the leaders in your organization treating AI as an operational redesign — or still as an innovation experiment?