Digital transformation 2026: What’s in, what’s out

I remind CIOs, “You will always be transforming.” Every two years, new business drivers emerge, such as the pandemic from 2020-2022 and automation-driven efficiencies from 2023-2024. We’re now in the gen AI era, where most CIOs are under pressure to shift from driving broad experiments to delivering business value and ROI.

As a result, CIOs need to refocus their strategies and communicate an updated vision for transformation. My 2025 article on what’s in and out for digital transformation stressed the importance of developing transformational leaders and AI-ready employees while avoiding AI moonshots and ending lift-and-shift cloud migrations.

In 2026, experts suggest that CIOs must transform IT, transition AI to customer experience (CX) opportunities, and double down on data governance and security.   

In: Reengineering IT’s digital operating model

In 2025, I wrote about how AI is the end of IT as we know it and how CIOs are rethinking IT for the agentic AI era. World-class IT organizations are setting higher expectations, partnering with departments on AI change management, and committing to lifelong learning.

With all the AI innovations impacting IT, CIOs will need to refocus their digital operating models to deliver more capabilities faster, at lower cost, and with higher resiliency.

Sesh Tirumala, CIO at Western Digital, says, “Velocity gets us ahead, resilience keeps us steady, and adaptability ensures we stay ahead. Direction matters, and in 2026, velocity is the real currency of success.”

How can CIOs aim higher when CEOs and boards are demanding ROI from AI? Jay Upchurch, CIO at SAS, says the best and brightest CIOs will snap up commercial responsibilities. “Top CIOs will sell customers and their divisional peers on technology like CMOs, and answer the constant call to do more with less like CFOs.”

I expect many CIOs will reorganize IT in 2026. Some will be mandated to reduce costs and headcount, while others will drive efficient collaboration in their product management, agile, and DevOps practices. Top CIOs will seek opportunities to guide reorganization across the enterprise as agentic AI creates new workflow patterns and cross-department collaboration opportunities.

“CEOs will conclude that AI adoption is no longer a technology problem but a workforce and management problem,” says Florian Douetteau, co-founder and CEO of Dataiku. “Instead of selling cloud migrations and data platforms, consultants will start selling organizational rewiring to prepare for AI-run operations. This shift creates tension inside enterprises because it surfaces the real blocker: leadership culture, not technology.”

Raja Roy, senior managing partner in the office of technology excellence at Concentrix, adds, “The new priority: operating models that support rapid learning, collaboration, and real-time evolution, keeping the human/AI balance aligned to the right tasks, whether an interaction calls for a human touch or machine efficiency.”

Recommendation: CIOs should review IT’s structure and agile practices to increase the effectiveness of delivering AI innovations and improve operational resilience.

Out: Underinvesting in data governance

Data governance is a critical function in global regulated enterprises, where governance, risk, and compliance (GRC) are critical top-down mandates. Midsize organizations are catching up, as they evolve to data-driven organizations and centralize data for AI initiatives.

While governing relational databases and warehouses is a relatively mature process, deploying agentic AI capabilities requires new tools and practices to extend data governance to unstructured data sources. 

“Unstructured data now moves too fast for manual oversight, and organizations can finally govern it as it’s created instead of cleaning it up later,” says Felix Van de Maele, CEO of Collibra. “In 2026, human judgment still matters, but AI-assisted systems, not spreadsheets or static controls, will carry the day-to-day load.”

Van de Maele suggests that AI-powered metadata generation for unstructured data, with integrated data practices for building reliable AI at scale is in, while CIOs should move away from manual tagging, siloed datasets, and one-time compliance efforts.

Additionally, many data governance leaders must get more granular controls on who gets access to what data. Authorizing users to full datasets and file systems is no longer sufficient as more organizations deploy AI agents on top of whatever data an employee can access.  

“Many organizations do not know where their sensitive data lives, who can access it, or how much is exposed across cloud and SaaS systems,” says Yair Cohen, co-founder and VP of product at Sentra. “Leaders in 2026 will treat governance as an engineering practice by embedding classification, tagging, and access rules directly into data pipelines, warehouses, and AI workflows.”

Recommendation: CIOs should be paranoid about data risks, take a sponsorship role in data governance, and ensure that improving data quality is prioritized in every AI initiative.

In: Targeting AI for growth and UX

In 2025, I warned CIOs about promoting AI as a driver of productivity and efficiency. Eventually, the CFO wants to see ROI, and this is one reason we saw significant technology layoffs in 2025.

I compiled over 50 expert predictions around 2026 on AI, from agentic workflows improving operations through gen AI embedded in customer experiences. I believe AI will have its Uber and Airbnb moment in 2026, as startups revolutionize customer experiences and disrupt slower-moving business-to-consumer (B2C) enterprises.

One easy way to embrace AI-enabled customer experiences is to upgrade call centers and chatbots without major infrastructure investments. Rob Scudiere, CTO at Verint, says, “Brands can layer an AI-powered chatbot onto their existing application instead of replacing an outdated telephony system and interactive voice response (IVR).”

When considering improving customer experiences, Pasquale DeMaio, VP of Amazon Connect, says to embrace systems that leverage AI and human strengths. “In customer support, agentic AI will manage routine requests while human agents will address complex issues with empathy and nuance, guided by AI insights and recommendations.”

CIOs should recognize a paradigm shift in UX, as data entry forms, customer journeys, and prescriptive reports get replaced with agentic AI capabilities. Focusing on AI in customer support is an easy entry point, as the entire customer experience, especially in ecommerce and SaaS tools, requires redesigning with AI capabilities.

“AI agents will become the frontend of the company as the primary starting point for any and all external contact,” says Antoine Nasr, head of AI at Forethought. “End-users will no longer have to try and navigate to the correct department and tool to get the help or information they need — they will simply interact with the company’s public AI agent in natural language. With that, agent design will become a key concern for several functions, not just customer support.”

Recommendation: Product-based IT organizations are a step ahead in anticipating how AI will evolve CX, and they should plan to segment and learn from early AI adopters.  

Out: AI experimentation without paths to short-term business value

Several research reports in 2025 highlighted how few AI experiments are being deployed into production and delivering business value. CEOs and boards will demand that CIOs narrow the portfolio of AI experiments and have real plans to deliver ROI from AI investments.

Conal Gallagher, CISO and CIO at Flexera, says in the next era of AI, execution matters more than experimentation. “CIOs will only continue to face bigger challenges and pressure to move beyond the AI experimentation phase and deliver clear, actionable, and measurable business outcomes.”

AI agents from top enterprise SaaS and security companies follow common patterns. These AI agents focus on a primary employee workflow, connect to multiple data sources, and aim to do more than complete tasks. CIOs will have to demonstrate the business value of how these AI agents guide employees in making smarter, faster decisions and the financial impacts of AI-revolutionized workflows.

“Agentic AI delivers measurable ROI in months, not years, because it replaces entire processes, not just parts of them,” says Luke Norris, co-founder and CEO of KamiwazaAI. “Each successful deployment accelerates the next, creating a self-funding innovation loop. More and more enterprises will be realizing this compounding ROI in the coming 6-12 months.”

Experts offer guidance on transitioning from an experimental to an outcome-based mindset. Kerry Brown, transformation evangelist at Celonis, says after years of big AI investment, it’s time to rethink end-to-end processes rather than just adding more automation on top.

“Leaders need to empower employees with visibility into how work really happens, and give them ownership in redesigning it,” says Brown. “When teams have that context and agency, they become true drivers of transformation and help create a faster, more direct path to ROI.”

Ed Frederici, CTO at Appfire, adds, “What’s out in 2026 is treating AI as a standalone, isolated initiative, and the next wave of digital transformation moves beyond scattered pilots to full operational integration. CIOs will treat AI as core business infrastructure rather than a special project — holding it to the same expectations for accuracy, security, and performance as every other critical system.”

Recommendation: Organizations with too many independently running AI experiments should revisit their AI governance strategy, communicate clear objectives, and prioritize where to build AI delivery plans. 

In: Implementing security before AI deployments

Nearly every transformational technology started with a gold rush to deliver innovations, and bolting on security afterward. CIOs will face pressure to move last year’s AI experiments into production this year, and we’ll have to see to what extent security will be implemented in initial deployments.

Many experts chimed in on where CIO’s need to get ahead of the curve. Here are three recommendations:

• Implement agentic AI observability and trust verification frameworks. “2026 marks a major shift in the threat landscape as agentic commerce takes hold, and in turn, AI-driven deception accelerates,” says Gavin Reid, CISO at HUMAN Security. “CIOs need visibility into how and what AI agents operate across their environments and deploy trust verification frameworks that continuously validate identity, intent, and behavior in real-time.”
• Establish security by design, especially around identity. “A unified identity layer is now a prerequisite for effective AI security implementation and is an urgent priority for any organization making AI investments,” says Ev Kontsevoy, CEO and co-founder of Teleport. “Organizations that embed these secure-by-design practices across development, delivery, and operations, and treat infrastructure security as a necessary mandate, will be best prepared for the transformational changes that AI will introduce.”
• Extend data loss prevention to AI-powered browsers. “AI-powered browsers like OpenAI’s Atlas and Perplexity’s Comet are one of the biggest blind spots in enterprise security,” said Rohan Sathe, co-founder and CEO at Nightfall. “Employees use them to research deals, draft customer outreach, and summarize strategy docs, giving agents with memory and sync direct access to logged-in Gmail, CRM, and code repos. Legacy data loss prevention cannot see this, since it was built for files, not browser-level activity, prompts, or clipboard moves.”

Recommendation: CIOs must partner with CISOs, legal, and risk management to clearly define AI security non-negotiables, platforms, and implementation requirements.

CIOs should expect the unexpected in 2026, whether driven by volatile economic conditions, new AI capabilities, or headline-making security incidents. My back-to-basics recommendations for digital transformation in 2026 aim to guide CIOs toward growth opportunities while improving operational resiliency.