Agent-based AI evolution and its impact on workforce and governance
AI agents are growing up. In 2025, many companies took early steps, some useful, some just hype. Now in 2026, we’re seeing the real shift begin.
AI agents are moving beyond simple tools. They’re quickly becoming intelligent systems that can manage entire workflows themselves, without needing a human in the loop. That’s a big shift. Craig Le Clair, VP and Principal Analyst at Forrester, put it clearly: “The AI agent is becoming the orchestrator of the process.” If you’re still designing systems with humans in the center as middleware, you’re behind. Humans are now endpoints. AI agents are orchestrating.
As these agents evolve, they don’t come neatly packaged. You get “agent sprawl.” That’s what Tom Coshow at Gartner calls it. Instead of one solution for everything, companies run dozens or hundreds of narrowly focused AI agents. They handle specific tasks, feed into each other, and collectively run multistep processes. This complexity makes management harder, but it also unlocks huge performance gains if handled right.
The flood of AI agents is already pushing CIOs into new territory. Governance, monitoring, security, cost, these won’t manage themselves. Agent management platforms are the next step. You’ll need systems that track dozens of agents, handle observability, offer cost transparency, manage marketplaces, and adapt to expansion. CIOs who ignore this now will spend more time playing catch-up later.
With this shift, the workforce must adapt. New tools mean new skills. Some existing roles will fall away. Others will evolve. And entirely new ones will emerge. Waiting for HR to solve this will slow you down. According to Forrester’s Le Clair, CIOs must lead the creation of a “skills progression roadmap.” You need visibility on which skills are rising, which are fading, and how fast you’re transitioning from human-run to AI-run processes.
If you’re a business leader, don’t delegate this to a taskforce. Own it. Get directly involved in aligning your technology teams and workforce planning. Agentic AI is not a trend, it’s architecture for how your business will run in the next few years. The vendors aren’t there yet. No one’s offering truly self-directed AI systems. But the components are coming together fast. The companies prepared for that convergence will see massive gains in speed, cost-efficiency, and scale.
Ignore the noise. Track the evolution. Get your workforce roadmap in place. Build governance that scales. And prepare for AI agents to become the drive train, not the accessories, within your enterprise systems.
Acceleration of upskilling and workforce transformation for AI readiness
AI is changing how companies operate, but the biggest barrier isn’t the technology, it’s people who aren’t ready to use it. In 2025, many businesses focused heavily on getting the infrastructure and tools in place. That was necessary. But now, in 2026, the real challenge is making sure teams actually know how to work with those tools, and use them productively.
The shift is visible. Jessica Hardeman, Global Head of Talent Attraction at Indeed, said it directly: “AI has made continuous learning non-negotiable.” That’s not just an HR talking point. It’s a business imperative. Organizations that don’t embed learning into the core of their AI strategy will waste time and resources on unused capabilities. Talent without training leads to systems without value.
And right now, most employees aren’t getting the training they need. According to a Jobs for the Future survey, only one in three workers say they’ve been offered any training in AI. More than half reported feeling unprepared. That’s a gap large enough to limit execution across your entire technology roadmap. If your AI tools aren’t being used confidently, from daily tasks to strategic decisions, you won’t see a return on your AI investments.
Some companies aren’t waiting around. Citi already required 180,000 employees to complete AI prompt training. They didn’t just encourage it, they made it mandatory. Walmart is working with OpenAI to launch an AI certification program. Google set up a $75 million fund to push AI literacy and created an AI fundamentals program. These are sharp moves. Well-timed upskilling isn’t a cost center. It’s an acceleration strategy.
Audi Rowe, leader for consulting transformation and digital strategy at EY Americas, is seeing it firsthand. CIOs and executives can’t afford to approach upskilling as a “nice to have.” She put it clearly: “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.” This includes both tech teams and non-technical professionals who’ll be using AI in their daily roles.
For business leaders, the takeaway is clear. You can roll out the best AI tools available, but if your teams don’t have the skills to use them creatively and intelligently, you’ll hit limits fast. Upskilling must be specific, measurable, and tied to value delivery. Make it part of your operational planning, not a separate initiative that fights for priority.
CIOs and executives should work together with talent leaders to invest in outcome-driven training, not just general awareness. Start with business outcomes, and design learning around that. If you’re serious about scaling with AI, then treating training like an infrastructure investment, not a side program, is the only move that makes sense.
Expanding state-level AI regulations amid federal challenges
AI regulation is no longer theoretical. It’s operational, and it’s already disrupting how companies build, launch, and govern their tech. In 2026, smart business leaders will need more than a compliance officer. They’ll need strategic visibility across federal and state-level rules that are starting to conflict.
The federal government, under President Donald Trump, issued an executive order in 2025 aiming to challenge state-level AI laws. This included the creation of an AI Litigation Task Force, which will target laws branded as “unconstitutional.” However, there’s lack of clarity around which laws qualify, and no court interpretations yet exist. According to Forrester Principal Analyst Alla Valente, it’s still unclear how the administration will define “restrictive and burdensome,” which creates legal uncertainty for companies operating across multiple states.
Despite this federal pushback, states are not slowing down. New York passed legislation in December that requires AI developers to disclose safety protocols publicly and report any safety incident within 72 hours. Texas enacted the Responsible AI Governance Act, which prohibits using AI for behavioral manipulation, infringement of constitutional rights, or generating illegal deepfakes. California implemented the Transparency in Frontier Artificial Intelligence Act, and Colorado’s AI Act, which regulates high-risk AI systems, is set to take effect by mid-year.
The result is a patchwork of fast-moving laws, each with different demands around transparency, auditability, and data use. For enterprise leaders, this creates operational friction. CIOs are now responsible for helping their organizations prepare comprehensive compliance frameworks that work across jurisdictions, because simply waiting to see how federal action plays out is a risk. You won’t be protected by inaction.
Lily Li, founder of Metaverse Law, draws attention to adjacent legal frameworks, like California’s regulations on automated decision-making technologies (ADMT). These demand that businesses disclose how they use personal data to automate decisions and require risk assessments and frequent updates to consumer notices. These types of adjacent rules often go unnoticed in federal debates, but they directly affect how your systems must be designed and documented.
For executive teams, especially those overseeing compliance-critical functions like finance, healthcare, or consumer data, there’s a simple message: Treat observability and risk reporting as core capabilities, not compliance afterthoughts. Systems that aren’t transparent enough to be explained to regulators will come with high legal and reputational risk.
As lawmaking continues in parallel at state and federal levels, companies must get ahead of these regulatory shifts, by building processes that can adapt, scale, and report clearly. The CIO plays a central role here. Don’t delay operational readiness until the federal vs. state battle is settled. Start building compliance into your AI systems now, because regulators already are.
Increased pressure on CIOs to demonstrate ROI from AI and data initiatives
The role of the CIO hasn’t changed. What’s changed is the urgency and volume of what’s expected. In 2026, tech leaders are under clear pressure, to deliver fast, deliver value, and prove their decisions are improving the business. Speed alone isn’t enough. What matters now is strategic alignment and measurable results.
This is where a lot of enterprises are getting stuck. They’re deploying AI tools but don’t see the value they expected. Large investments sit underused or tangled in integration. CIOs must now lead the charge in connecting all of it, data, tools, processes, training, into direct business outcomes. Christie Struckman, Distinguished VP Analyst at Gartner, described it well: “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.”
To act effectively, CIOs need to shift mindset from execution to orchestration. The job isn’t just standing up platforms or enabling automation. It’s understanding the business strategy, identifying which capabilities will create differentiation, and focusing resources there. Everything else takes a back seat. Budget environments are tightening. There isn’t room to fund every initiative.
Martha Heller, CEO of Heller Search, sees this moment as one of decisive change. For years, many enterprises collected large amounts of data without unlocking much value. But the introduction of tools like ChatGPT made the power of usable data obvious. Now, boards and executives expect data strategy to produce real business insight, not just dashboards. “To automate processes, to produce data, to produce value… that has always been the job,” she said. The difference now is that no further justification is needed. The expectation is already there.
Data sprawl is real. Enterprises have accumulated large, fragmented datasets, most of which are underleveraged. This is the year that has to change. The CIO is expected to operationalize data, not just govern it. That means enabling AI systems that can learn from data and drive actions across the business. It also means making that data accessible, reliable, and relevant, without adding new layers of complexity.
There’s no advantage in being hesitant. If your data is disorganized and your systems don’t talk to each other, you’ll spend the year solving legacy problems while competitors accelerate. As Heller put it clearly: “If you don’t get your data right… I think we’re going to see the have and have-not divide.” She encourages CIOs to look at what worked during digital transformation years ago and treat AI as the next strategic shift.
In summary, enterprise technology has never offered more possibility, but the window for proving value is shortening. Executives are looking for momentum, not just roadmaps. Pilot projects need to scale. Metrics need to connect directly to revenue or operational gains. And CIOs must own the narrative that ties data, AI, and productivity together. This isn’t about defense. It’s about owning the shift and showing results that justify continued investment.
Key takeaways for leaders
- Manage agent sprawl before it disrupts operations: CIOs should prepare for a surge in AI agents by investing in platforms that enable orchestration, governance, and visibility. Establishing a clear roadmap for workforce reskilling is essential as AI agents take over process orchestration roles from humans.
- Connect upskilling to measurable business outcomes: With most employees still untrained in AI use, leaders must embed targeted learning into transformation plans. Training should be tied to productivity gains and not siloed in HR to avoid wasted tech investment.
- Treat AI regulation as a multi-jurisdictional compliance priority: State-level AI laws are advancing regardless of federal pushback, creating a shifting legal environment. CIOs should build flexible compliance strategies that accommodate both direct AI rules and adjacent data-use regulations.
- Make data and ROI delivery core to CIO strategy in 2026: Technology leaders must focus on initiatives that clearly demonstrate measurable business value from AI and data investments. CIOs should prioritize unifying fragmented data estates and translating them into real outcomes that resonate with executive teams.
