Industry executives and experts share their predictions for 2026. Read them in this 18th annual VMblog.com series exclusive.
By Matthew Hausknecht, Principal AI Researcher and Alex Conway, Principal Software Engineer at DataRobot
Organizations today face a variety of challenges when deploying AI workflows at scale. In 2026, we foresee a renewed focus on deployment (instead of AI experimentation) to drive meaningful returns on AI investment (ROAI).
Forward looking organizations will move beyond the evaluation of models in isolated environments and will begin investing heavily in governance, unified tooling, and cost control to ensure they’re able to deliver AI that works for the business and its customers.
The predictions below from myself and my colleague, Alex Conway, provide insight into several ways we foresee AI evolving in 2026.
AI makes a much bigger impact in customer service
“In the coming year, we expect to see AI workflows in customer service environments increasingly move from pilots into production.
Many organizations have been hard at work �getting their data right,’ and in 2026, early adopters will begin deploying their AI workflows at scale. As a result, human customer service agents will focus on only the most sensitive requests while moving into ‘AI trainer’ roles, ensuring AI agents are operating effectively and staying apprised of the latest product updates, company messaging, and more.
As AI agents become more effective and reliable, human workers will play an essential role in ensuring they are representing the brand faithfully, at every turn.” – Alex Conway
Physical AI takes a major step forward
“In 2026, we’ll see AI-powered robots begin to graduate from controlled environments to more unpredictable real-world scenarios.
While many robots are partially operated by a human, we expect they’ll continue pushing towards greater autonomy as they learn from the messy, real world. They’ll fail early and often, but just like with humans, failure is often the greatest teacher – and we know that AI is an incredibly quick learner.
2026 will be physical AI’s breakthrough moment, with robots going from proof-of-concept demonstrations to economically viable realities in a variety of controlled and semi-controlled environments.” – Matthew Hausknecht
It’s clear that 2026 will be defined by a critical pivot for organizations that are embracing AI. However, the challenge of moving pilots to production hinges on an organization’s focus on effective deployment processes, especially when dealing with critical use-cases like customer service and physical environments.
It will be vital to deploy AI with governance and observability embedded from the beginning – creating a foundation of trust between human and digital coworkers that will be essential to the success of our agentic future.
The ROAI generated in 2026 will not only drive the next wave of innovation but will help businesses realize the true potential of AI in the workplace and beyond.
##
