International Business Machines A Guide to IBM Software: Upgrading to Overweight with Greater Confidence in Software Acceleration Ahead

Research evidence excerpt

International Business Machines A Guide to IBM Software: Upgrading to Overweight with Greater Confidence in Software Acceleration Ahead

Brian Essex, CFA AC North America Equity Research

(1-212) 622-5990 23 June 2026 J P M O R G A N

brian.essex@jpmchase.com

Secular Drivers

We view container adoption as a primary structural tailwind, and AI accelerates it.

Containers isolate applications and their dependencies, so teams can fix errors without

taking down an entire system, which we view as an effective way to manage and secure

AI deployments. If an AI agent breaks out of its environment, IT teams can also kill a

container with minimal disruption to surrounding processes, which we see as a clear

security advantage over virtual machines that require booting a full virtualized OS. We

expect AI to accelerate the shift from VMs to containers, and the data supports that

view. McKinsey projects 95% of businesses will run containerized applications by 2029,

CNCF data shows 84% of enterprises already implementing containerization, and 75%

of AI-related deployments will use containers by 2027.

Containers also enable an architectural shift we think AI reinforces, from monolithic

application design toward modular, microservices-based design. AI relies on

collaborative, specialized agents, which lets enterprises update individual components

without retraining a monolithic model, while microservices let functions like AI analysis

or data ingestion scale independently. Generative AI tools break down legacy code,

document service interactions, and generate code for new APIs and containers. Modular

systems carry higher fault tolerance because one component’s failure does not bring

down the whole application. These applications still require a system of record.