In 2025, CIOs spent countless hours shaping AI adoption playbooks, keeping tabs on regulation shifts and reviewing the latest offerings from a fast-evolving vendor market.
While AI owned the enterprise main stage last year, lingering questions about the technology's rapid development are landing right back on the CIO desk in 2026, as new capabilities come online and organizations look closely at the ROI metrics of previous projects.
The new year welcomes C-suite leaders with a string of trends to watch.
Executive board members will ask tech chiefs how they plan to manage the growing army of agentic AI tools, how to advance ROI and how AI regulations will affect operations.
Amid the changes, CIOs themselves are grappling with the evolution of their own role and how training efforts should change in response to AI's swift adoption.
Here are four trends CIOs should keep an eye on this year:
1. Agents bring sprawl, spark skills assessment
Agentic AI was all the rage in 2025, and experts expect enterprise interest and investment to continue — bringing with it confusion, complexity and promise.
A sticking point will be semantics, as CIOs continue to separate bloated vendor promises from technology that can deliver. One rule of thumb? No vendor has a truly agentic system on the market today, according to Forrester VP and Principal Analyst Craig Le Clair. AI agents, on the other hand, will mature in 2026.
It’s another way of saying agentic AI is a progression, where systems that can create new tools, establish new processes and resolve issues during runtime is an end goal. AI agents are part of that progression. In 2025, agents tackled processes designed for humans while humans acted as orchestrators at the center. That will change in 2026, with humans operating more as endpoints than middleware, according to Le Clair.
“They're moving humans out of that middle position, and the AI agent is becoming the orchestrator of the process,” he said.
The shift will have an effect on enterprise skills, Le Clair added. To prepare, he advised that enterprises establish a skills progression roadmap that reflects how quickly they plan to move from human-driven to agent-driven processes. The roadmap could consider skills that are emerging, critical for today but declining in importance and those already made obsolete by AI. IT leaders will play a pivotal role in building out these assessments.
“The CIO has to collaborate with the business to really create a progression for a growing agent capability,” Le Clair said, adding that CIOs need to help the business “decide what is the progression of that agent, what is the agent's capability, how is that progressing over time.”
Not only will AI agents become more mature in 2026, there will be more of them, creating what Gartner Senior Director Analyst Tom Coshow referred to as “agent sprawl.”
In 2025, enterprises realized that to get reliable results, AI agents needed discreet tasks and proper data to be successful, Coshow said. Narrowly defined agents will work together to complete multi-agent workflows, but they will also create a management headache for CIOs and pave the way for agent management platforms.
“Some people call it orchestration. Some people call it observability and governance,” he said. “I call it an agent management platform.” These should include libraries of built agents and features for marketplace management, observability, cost reporting, security and associated tooling.
“Everybody is working on some version of this right now, from hyperscalers to startups and all the SaaS platforms in between,” Coshow said.
One more AI agent prediction for 2026? “Somebody is going to build a voice AI that blows everybody's mind,” he said.
2. Upskilling efforts accelerate in search of enterprise value
In 2025, leaders spent countless hours thinking through what worked and what didn't amid AI scaling efforts. Although technical and governance challenges loomed, the trickiest component in deployment plans is often related to people.
As workers become more familiar with the AI capabilities they have on hand, reskilling efforts are critical to deliver AI's promised boosts in productivity. In 2026, the right training efforts will help determine which businesses enjoy successful AI adoption plans — and which will waste resources in underused tools.
"AI has made continuous learning non-negotiable," Jessica Hardeman, global head of talent attraction at Indeed, said during a virtual media roundtable in November. "Instead of focusing only on deep technical specialization, we're trying to build learning paths that help every employee understand how to apply AI into their day-to-day work."
Despite widespread AI initiatives across enterprises, employee training is lagging behind, according to a Jobs for the Future survey. Just one-third of workers said their employers offered training in AI, and more than half reported feeling unprepared to use the technology.
A push to reskill workers is already underway. Citi required 180,000 of its staffers to complete AI prompt training last year as part of broader AI rollout plans. Retail giant Walmart tapped OpenAI to stand up an AI certification program set to begin this year. And Google unveiled a $75 million AI training fund last year alongside an AI fundamentals program.
In 2026, CIOs will spend more time aligning upskilling efforts with clear value delivery, said Audi Rowe, EY Americas leader for consulting transformation, digital strategy and experience.
“There has to be an upskilling of the workforce to understand how to shift into roles that are higher value or more focused on innovation,” Rowe told CIO Dive. “I believe that there's going to need to be quite a steep upskilling, both within a CIO organization as well as more broadly.”
3. States continue AI regulation efforts despite executive order
Despite pushback on AI regulations from the federal government and Big Tech, enterprises shouldn’t slim down their compliance frameworks as states continue to pass laws governing the technology.
Efforts to stem the tide of state-led AI regulation in 2025 culminated in an executive order from President Donald Trump that will likely begin to play out in 2026. Trump mandated the creation of an AI Litigation Task Force to target state AI laws deemed unconstitutional. What those laws are remains to be seen, said Forrester Principal Analyst Alla Valente.
“How is he going to define what is restrictive and burdensome?” Valente said of the executive order. “It’s unclear which state laws will be challenged.”
The threat the executive order poses to states’ ability to regulate the technology hasn’t stopped U.S. states from passing and adopting AI laws.
New York Governor Kathy Hochul signed legislation in December that established safety and transparency frameworks for large AI developers, requiring the companies to publicly publish information about their safety protocols. The law also requires frontier AI developers to report AI model safety incidents to the state within 72 hours of discovering the incident.
Across the country, the Texas Responsible AI Governance Act went into effect Jan. 1. The law prohibits creation and use of AI systems for purposes such as behavioral manipulation, infringement of constitutional rights and distribution of unlawful deepfakes.
While New York and Texas are the latest to pass AI laws, they’re not alone. California’s Transparency in Frontier Artificial Intelligence Act — which New York modeled its law after — took effect Jan. 1, while the Colorado AI Act regulating high-risk AI systems is expected to take effect June 30, 2026.
Even at the federal level, there’s bipartisan consensus that AI systems should be evaluated. Sens. Josh Hawley, R-MO, and Richard Blumenthal, D-CT, introduced the Artificial Intelligence Risk Evaluation Act in September 2025 to create an advanced AI evaluation program that would mandate frontier AI model developers’ participation.
Heading into 2026, businesses that develop and deploy AI systems will need to pay attention not just to new AI laws and legislation, but AI-adjacent regulations such as California’s automated decision-making technology regulations because it’s unclear which laws will fall under the scope of Trump’s executive order, said Lily Li, founder of Metaverse Law.
“They govern how companies process personal information in order to make automated decisions about consumers,” Li said about Califiornia’s ADMT regulations. “Those require risk assessments, disclosures and a lot of updated notice requirements.”
Some of the challenges enterprises will face with broader state AI laws such as the Colorado AI Act include understanding the scope of AI system transparency and documentation requirements, and determining which risk assessments to conduct internally and report to state regulators, Li said.
While these are some of the areas the Trump administration might focus on, it remains to be seen how any challenges will affect states’ implementation of their AI laws, Li added. Enterprises need to keep a handle on what data they’re using to train AI models, as well as understand the logic and parameters of their AI systems, as laws are passed that will eventually require businesses to disclose that information.
“The risk assessment and reporting requirements to state regulators will be under the most scrutiny by the Trump administration,” Li said. “[But] there’s actually a lot more variability there in terms of how successful the challenges will be.”
4. CIOs face greater pressure to produce ROI
The role of the CIO will not see significant transformation in 2026, but the pressure to produce has never been higher and the opportunities to effect change have never been greater.
“If I had to summarize what I think the challenge is going to be for CIOs, it's figuring out how to survive and where to thrive,” said Christie Struckman, distinguished VP analyst at Gartner.
CIOs will be faced with high-volume technology demands and expectations that the tools are delivered faster, at a higher quality and provide an ROI, Struckman said. To survive, CIOs will have to figure out how their IT departments can meet those demands and make difficult, sometimes painful, decisions quickly to get there, she said.
To thrive, they should start by assessing what the business needs. “That is the only starting point: What is your enterprise trying to do? How fast are they trying to get there? And what do they think is going to be the differentiator?” Struckman said.
The demand is too high and the budget constraints too tight to do it all, Struckman said, so CIOs should look for investments that can show value and experiment with pilot projects.
To Heller Search CEO Martha Heller, 2026 might be the year when data finally gets its due.
After years of acquisition and businesses doing just enough to remain profitable without a strong data foundation, companies are finding themselves at an inflection point: They have a ton of data and none of the value. CIOs will be expected to change that.
“The role of the CIO in 2026 will be to do the role of the CIO since it was invented,” Heller said. “To automate processes, to produce data, to produce value. That has always been the job.”
The difference in 2026: CIOs no longer have to make a case as to why data estates are critical to a company’s success. ChatGPT did that when it came on the scene in 2022.
What CIOs will be challenged with is donning the hat of the chief technology transformation officer where they bring together sprawling data landscapes, pave the way for upskilling employees to a new AI level of work and find ways to tell the story of data and AI to the executive team so that they understand and can prepare for the existential fork in the road, Heller said.
“If you don't get your data right, do you remember Radio Shack? Do you remember Blockbuster?” she said. “I think we're going to see the have and the have-not divide.”
She reminded CIOs they’ve experienced moments of transformation before and suggested they dust off their playbooks on how they integrated digital into the business as a reminder of what’s worked in the past. She encouraged they make the transition to AI a team sport. And she advised IT leaders to grab the opportunity in front of them.
“AI is the portal through which you can [become a business leader] because AI is tech, but it is absolutely a capability,” Heller said. “Go step through it.”
