Every January, MIT Sloan publishes a short outlook from its faculty on what is going to matter most in AI and work over the next twelve months. The 2026 edition is unusually concrete: less about model capabilities, more about the organizational changes leaders need to make to capture the value of those capabilities.
Themes the faculty are watching
- Productivity is plateauing in pilots, accelerating in operations. The companies still running disconnected proofs-of-concept are seeing diminishing returns; those embedding AI into core workflows are seeing compounding gains.
- Middle managers are the new bottleneck. The skill of designing the right prompt, guardrail and review cadence for a team of agents is in short supply.
- Job redesign is overtaking reskilling. Splitting old roles into "agent supervisor" and "exception handler" tracks is producing better results than retraining everyone to use AI tools.
What it means for marketing leaders
Marketing was one of the first functions to adopt generative AI and is now feeling the operational pain that comes with that. The 2026 question is no longer "should we use AI?" but "how do we redesign our team so the work flows through AI by default?".
The VietMarkAI take: the highest-impact 2026 investment for most marketing organizations is not another AI tool — it's a deliberate redesign of who does what, with an explicit place for agents in the org chart.