Automation

Despite widespread adoption efforts, fewer than one-quarter of enterprises have successfully scaled AI across multiple business units, according to Gartner data. The IT advisory firm surveyed more than 1,300 leaders from organizations with more than $50 million in revenue between January and April of this year. 

The lag in adoption isn't slowing plans to roll out the technology, with 85% of tech leaders planning to up their AI investments next year. About 11% of the organizations that responded said they lacked visibility into what they spent on AI in 2025, according to Gartner.

Low visibility into AI spending heightens the risk of missing ROI, Tina Nunno, distinguished VP and fellow at Gartner, said in the report. “Without disciplined measurement tied directly to business outcomes, organizations risk wasted resources and unmet expectations,” she said. 

Enterprises continue to fuel AI investments despite a lack of clear ROI, but this approach can muddy an organization’s chance at finding the right use cases for it. 

Nearly three-quarters of senior business executives said they’ve scaled fewer than 25% of their AI pilots successfully, and two-thirds said their organization struggles to measure the ROI generated by AI to prove the positive benefits executives tout, according to an August Infosys report.

Many enterprises aren’t as prepared for AI deployment — especially agentic systems — as they think. Only 1 in 5 senior managers and C-suite executives said their organization is prepared to redesign business processes to run autonomously with AI agents, an August Deloitte report found.

Enterprises that constantly track the ROI of their AI initiatives, treat it as a portfolio of value and regularly assess project performance see greater returns, Gartner’s data showed. Companies that follow this framework and discontinue underperforming initiatives reported positive returns in 81% of their AI initiatives. 

“We want companies to get more out of AI, rather than risk spending more and either not getting appropriate return on investment, or creating more technical risk,” Nunno said in an email to CIO Dive.

Tech leaders can fall prey to AI hype cycles, Nunno said. The most widely pursued AI use cases, such as cybersecurity, threat detection and IT service desk automation, don’t always deliver the highest ROI. 

Intelligent IT asset and cost optimization, synthetic data generation and automated code generation and refactoring were among the top three AI use cases with positive returns, according to Gartner.

However, the AI use cases that will prove the most value for a company are the ones that target specific business needs and build necessary data foundations when needed, Nunno said.

“CIOs should always prioritize use cases where the business case and financial logic make sense for their specific company, AI maturity and business needs,” Nunno said. “Having rigor both around the business case, and ensuring it meets your company’s strategic vision for AI will ensure CIOs choose the best use cases to maximize value.”

Enterprises are leaning on hybrid cloud strategies to accommodate growing AI workloads, a report from advisory firm Information Services Group found. 

More than 80% of enterprises are revisiting their cloud plans to better support AI, the report found, citing wishes to strengthen their operational resilience and improve consistency, cut costs and improve performance. Hybrid approaches that are gaining popularity often include a combination of public cloud, private cloud, colocation, edge and sovereign environments, or all of the above, the report found.

Organizations looking for flexibility and cost-consciousness are redesigning their operating models to meet the demands of AI, Anay Nawathe, ISG cloud delivery lead for the Americas, said in a statement. "Hybrid cloud has become less about where workloads run and more about how enterprises maintain control across increasingly diverse environments,” he said.

As enterprise demand for AI compute rises, and executives worry about increasing costs or vendor lock-in, more companies are taking a hybrid cloud approach. 

Preparing to expand AI use is one of the central reasons to change cloud strategies, ISG’s report found. Enterprises are investing in GPU-enabled architectures, distributed data platforms and hybrid AI operating models while adopting AI platforms. 

Organizations are seeking unified platforms that combine observability, automation and financial management into one operating framework, the report found. Performance and cost management were also high priorities for enterprises seeking hybrid cloud environments. 

Vendors in the sprawling cloud market have quickly moved to address the growing needs for AI-ready compute. Neocloud providers, which offer AI-optimized infrastructure, will earn 20% of the $267 billion AI cloud market by 2030, a recent Gartner study found. Apple opened up access to private cloud compute in June, while Blackstone and Google launched a compute-as-a-service offering in May. 

Increasing AI workloads and GPU-intensive infrastructure is driving up demand for cost transparency and optimization across private cloud, Kubernetes, edge and sovereign cloud environments, the report found. 

Sovereignty is a major factor behind diversifying cloud offerings, the report said, as enterprises want more control over infrastructure, data residency and AI governance for their long-term AI plans.

"The most effective hybrid platforms no longer treat operations, cyber recovery and cost management as separate disciplines," said Shashank Rajmane, ISG principal analyst and lead author of the report, in press release about the report. "Enterprises increasingly expect providers to unify these capabilities, improving business continuity and long-term performance." 

Sponsored content

By Broadcom

The landscape of artificial intelligence (AI) is rapidly evolving. Enterprise IT and data science teams are moving beyond the era of single-task AI models to a future dominated by agentic AI. Agentic AI refers to complex systems in which multiple intelligent agents collaborate to achieve sophisticated goals.

Possessing varying degrees of autonomy, agents will interact, communicate, and make decisions. These agents may be organized in hierarchical structures, with higher-level agents overseeing the activities of subordinate agents. Through these advanced approaches, teams of AI agents will be able to undertake tasks such as designing new products, managing complex supply chains, and even optimizing existing process workflows that have been built over decades.

A Paradigm Shift Creates Challenges

This paradigm shift, while promising immense potential, introduces unprecedented challenges in terms of management and control. The potential benefits of agentic AI risk being overshadowed by increasing complexity and even chaos. Without a centralized orchestration and event management solution, agentic AI systems can quickly spiral, becoming unmanageable and unpredictable, which can pose several significant consequences:

Workflow breakdown. Without a proven system to manage dependencies and task execution order, workflows can easily break down, leading to delays, errors, and missed objectives. 

Resource conflicts. Agents might compete for the same resources, resulting in inefficiencies, bottlenecks, and system instability.

Communication chaos. Unstructured and uncoordinated communication between agents can lead to inconsistencies and errors. 

Loss of visibility and control. Without a centralized monitoring system, it becomes difficult to understand what agents are doing, how they are performing, and where problems are occurring. This makes it challenging to diagnose issues and take corrective action.

Security vulnerabilities. Systems will be vulnerable to cyberattacks, data breaches, and other risks without proper access control, security policies, and audit capabilities. 

Ethical concerns. Without solid governance and oversight, it becomes challenging to ensure that the agents are operating ethically and responsibly. This can lead to unintended consequences and potentially harmful outcomes.

Enterprise Orchestration Requirements

To maximize the potential of agentic AI, while mitigating its risks, teams must take a new approach to management and control. Managing a multitude of autonomous agents, ensuring their coordinated action, and maintaining control in dynamic environments requires a sophisticated orchestration framework. Teams must establish robust capabilities in the following areas:

Agent management. Defining, scheduling, and managing complex workflows involving multiple agents requires a centralized orchestration platform. This platform must be capable of handling dependencies between tasks, ensuring that agents execute their responsibilities in the correct sequence. The AI agent market is projected to grow from $5B to $47B in 2030 with a 44.8% CAGR. Agentic AI will increase event volumes by orders of magnitude, making it an absolute imperative that teams establish dedicated, centralized systems that enable highly efficient workflow management and control.

Resource allocation. Agents will require access to various resources, including computational power, data storage, and, in some cases, physical infrastructure. Orchestration is crucial for efficiently allocating and managing these resources, preventing conflicts, and optimizing utilization.

Communication and data flow management. Facilitating seamless communication and data exchange between agents is essential. However, these environments will create a complex web of connectivity. For example, with just 100 agentic AI agents, approximately 5,000 potential communication pathways might be required. Automation teams require an orchestration platform that provides the necessary infrastructure for message routing, data synchronization, and ensuring data consistency across the system.

Monitoring and control. Maintaining visibility into the activities of numerous autonomous agents is critical for ensuring the system’s overall health and performance. However, given their high churn rates and frequent training iterations, these environments will generate trillions of data points daily. Consequently, groups require a robust event management system, integrated with the orchestration platform, that enables real-time monitoring of agent activity at a massive scale. These capabilities are essential in enabling early detection of anomalies, debugging and prompt intervention.

Governance and security. As agents become more autonomous, governance and security become paramount. Automation teams require an orchestration platform that enforces policies, manages access control, and ensures agents operate within ethical and legal boundaries. This is particularly critical for higher-level agents that wield significant control over the system. The growth is supported by market analysis which indicates 46.6% CAGR in spend for AI Governance to $132B by 2030.

Scalability. As the number of agents and the complexity of tasks increase, groups need an orchestration solution that can scale accordingly. Gartner predicts 33% of enterprise software applications will include agentic AI, up from less than 1% in 2024. Consequently, there is a need for an orchestration platform that can accommodate the growing volume of events, data, and communication, without performance degradation.

Agentic AI Is Here: What Benefits Does Your Enterprise Automation Solution Provide?

With the capabilities of modern enterprise automation platforms, such as those provided by Automation by Broadcom, enterprises can maximize the enormous potential promised by agentic AI and get the following powerful benefits:

Boost operational and cost efficiency. Harness centralized, unified orchestration, control, and visibility, streamlining implementations and ongoing administration. Gain the intuitive, timely insights that enable teams to intelligently manage workloads and resources.

Optimize service levels and reliability. Leverage a single platform for defining, scheduling, and monitoring complex workflows and SLAs. Gain the capabilities needed to detect anomalies early and quickly address emerging issues.

Ensure continuous compliance. Gain the visibility and control needed to enforce policies, track and govern access, and ensure that agent operations are aligned with relevant corporate policies and regulations.

For more from Automation by Broadcom on AI-powered automation, watch our 2026 Automation Virtual Summit: The Next Horizon on-demand today!

Employees are less comfortable and confident with AI use than their managers, and the power dynamic between AI decision-makers and those instructed to use the technology will likely continue that trend, a TeamViewer report found. The remote access software company worked with Sapio Research to survey 4,200 managers and employees globally this spring.

Although three-quarters of respondents — both employees and managers — said they use AI at least once a day, 61% said they’d prefer it if AI had no independent action abilities.

More than half of respondents verify AI outputs often or always before using them, and about the same proportion said they don’t always know when to trust AI and when to verify it, according to the report.

Organizations must work toward trusted autonomy for their AI systems instead of blanketed automation, said Oliver Steil, CEO at TeamViewer, in a press release. “AI is already changing how people work, but the next phase will be defined by trust,” he said.

Although managers and employees alike are getting more comfortable using AI in their workflows, those leading adoption efforts generally feel better than frontline workers about AI and are more willing to delegate work to it. Employees reported that they feel greater personal risk from autonomous systems because of their lack of control over how and when AI is deployed. 

Software has traditionally been designed around what people tell it to do, but AI systems and agents change that model, Mei Dent, chief product and technology officer at TeamViewer, said in a press release. 

Insecurities around AI outputs aren’t causing workers to fully avoid the technology, but it might be limiting what they’ll use it for. AI is most frequently used once or twice a day, the survey found, and usually for tasks that help a user understand information, generate a starting point or complete a routine task — outputs that can be reviewed before they’re used. 

Only 11% of respondents said they’re comfortable with AI working autonomously, or taking action that they haven’t reviewed and approved.

Despite lagging trust in AI systems, enterprises are forging ahead on giving them autonomy. One-quarter of business and technology leaders said they have complete trust in their AI systems, a June survey published by Kyndryl found, but more than four-fifths said they expect autonomous AI agents to make decisions with material business impact within the next year.

Organizations struggling with trust often lack governance and accountability frameworks needed for AI outputs, Steil said. Respondents said they worry that AI systems could expose sensitive information, that it might not understand the context of the information to create the right outputs and that AI could act beyond its intended scope. About one-fifth of respondents said they’re concerned they will be held responsible for AI’s mistakes. 

“[Organizations] will succeed by building systems that operate safely, transparently and accountably within clearly defined guardrails,” Steil said. “The real opportunity is to use AI and automation to prevent disruption, improve digital experiences and free people to focus on higher-value work.”

Organizations must give their staff the confidence that autonomous systems are secure, explainable, reversible and governed by people, Steil said. 

A good starting point is establishing which actions are low-risk and pre-approved to be completed autonomously and which ones require human review. Employees should be involved in the process of determining autonomy to understand the authority a system has and where responsibility for errors lies. 

Organizations should also work toward ensuring that the risks of autonomous systems are distributed fairly across different workplace groups, rather than lying with one team or individual. 

“Autonomy will scale when organizations can prove it works in specific, trusted use cases, then expand it with confidence,” Dent said.

AI trust isn’t keeping pace with adoption, according to a Tether survey of over 5,000 AI users conducted by market research group Kantar. While 82% of respondents said they used more than one AI chatbot over the past year, nearly half reported they do not trust current AI companies.

Concerns about data privacy and collection, accuracy and reliability and dependence on internet connectivity ranked highest among respondents. Most users said AI companies should be more transparent about how they use and store data.

“We believe that people shouldn't have to choose between protecting their data or signing it over to gain access to the latest AI tools, and this research is an indicator that the market agrees,” Tether CEO Paolo Ardoino said in a press release. “The research also shows, however, that without comparable localized alternatives, users will willingly give up their data.”

AI is everywhere: in the workplace, on your smartphone and embedded in product interfaces. People use ChatGPT and Claude to solve problems, conduct research and even get personal advice, handing over their data in exchange for access to cloud-based AI tools.

Amid widespread AI adoption, data ownership is more essential than ever. More than half of the Tether survey respondents said they want to own data created through AI interactions, but only 35% were confident personal data was secure.

Despite the concerns, more than half of respondents reported they would trade some privacy for AI-enabled convenience or share personal data for lower-priced AI services.

Attitudes toward AI also vary widely depending on geographic location. In the U.S. and UK, users have greater skepticism toward AI and more concerns about health and medical data privacy, while users in India and Nigeria were more optimistic and had stronger interest in cross-device functionality.

The data shows people are using AI constantly while growing increasingly uncomfortable with cloud-based models, according to Ardoino, creating a clear opening for local AI that keeps data on-device.

“By building options for local, private AI tools that can match the speed and power of current market offerings, we can raise the ubiquity of AI technology while doing away with barriers that discourage new users,” Ardoino said. “The future of society can thrive only if AI becomes open and accessible to everyone.”

More than half of companies have only partial visibility into their data amid rising concerns over data oversight and transparency, according to a Thales report conducted by the Cloud Security Alliance. The report surveyed 210 IT and security professionals.

While three-quarters of organizations say they’re confident in their ability to secure unstructured data, 68% say a majority of their data remains unprotected.

In this climate, tech executives perceive AI as a solution as well as a threat. Nearly half of respondents view AI as a top future security risk against unstructured data, yet 40% see it as a key tool for managing it.

Unstructured data is making up a growing proportion of enterprise data, with nearly one-third of companies saying it accounted for more than half of their annual data growth.  

The report also found governance practices are struggling to keep pace with the rate of unstructured data growth. 

“The explosive growth of unstructured data — estimated by Gartner to account for between 70% and 90% of enterprise data—has become a defining characteristic of modern organizations,” said Hillary Baron, AVP of Research at CSA and lead author of the report, in a press release. “While it enables significant operational value, it also introduces substantial security risk.” 

Fragmented tooling is compounding the problem. Nearly one-third of organizations use more than 11 tools or more to manage unstructured data, creating operational silos that hinder holistic governance practices. Many organizations still rely on manual processes for classification and monitoring, an approach that struggles to keep pace with the volume of data growth.

In this landscape, AI is emerging as both a potential remedy and a new source of risk.

On one hand, enterprises are looking at the tech to automate processes such as data discovery, classification and threat detection. On the other, scaling AI tools without proper oversight runs the risk of greater vulnerabilities later down the line. Without clear visibility, AI models risk being trained on incomplete or biased datasets.

With only 9% of organizations able to scan data in real time – and nearly a quarter unable to scan at all – AI systems risk amplifying existing blind spots rather than resolving them.

“Without addressing these gaps, AI, automation, and emerging technologies will only increase exposure,” said Todd Moore, VP at Thales. 

AI is only as good as the data it’s trained on, with nearly 3 in 5 business leaders saying key decisions are made based on inaccurate or inconsistent data, according to a SoftServe report published last year. Moving forward without that foundation can lead to wasted resources, with many organizations stuck in the pilot phase as ROI lags and data challenges derail AI projects.

As businesses look to deploy AI agents more widely, AI ambitions hinge on a reliable data backbone, prompting organizations to prioritize data skills in upskilling efforts.

Foundational readiness will be the defining factor in whether unstructured data security efforts succeed, according to the Thales study. Organizations that invest in these areas — visibility, classification, governance, and scalable operations — will be better positioned to manage an increasingly complex data landscape.

AI automation is expected to force meaningful change in operational capabilities for companies as they shift from digital to autonomous businesses, 80% of CEOs said in a Gartner report. The report surveyed 469 CEOs and other executives worldwide in late 2025. 

Some CEOs said AI will negatively affect their profit models, as agents could disrupt existing intermediated systems or price negotiation processes. The shift could force executives to pursue recurring, outcome-based revenue models, the report said. 

The transition to autonomous businesses requires executives to prioritize how work gets done and how human value is delivered, Jennifer Carter, senior principal analyst at Gartner, told CIO Dive. They will be able to focus on the intrinsic, specialized value that employees can bring to their role. 

Enterprises are in the early stages of deploying autonomous AI at scale, but expect that they will ramp up their usage in the next several years. The long term effect on business models remains to be seen.

More than half of CEOs said automation was currently limited to specific tasks, but only 13% expect to remain at that level by the end of 2028. One-third of CEOs said their organizations will use self-learning and adaptable AI tools in human decision-making, and more than a quarter plan to use AI primarily without human intervention. 

“We're looking at our entire end-to-end operations and saying: How can we create entire workflows that, fundamentally, run themselves?” Carter said. “They can be managed by humans, but really turning inwards with the opportunity for automation.”

A recent Writer study found that tech leaders are feeling the pressure of AI adoption, with 61% reporting they fear losing their job if they fail to successfully lead their organization through AI deployment. Some say their own skills could become obsolete in the age of AI. 

Still, executives said they believe they can use AI to deepen relationships with existing customers and non-human, machine customers. Gartner expects enterprises will expand their use of dedicated business units for machine customers. 

The increased reliance on agents and autonomous workflows intensifies an enterprise’s need for trust, accuracy and data integrity, the report said. As adoption accelerates, the foundation on which companies deploy their AI becomes a core requirement, and poor AI hygiene will have lasting effects, Carter said. 

As CIOs take expanded responsibility over AI at their organizations, they must look inward at their operations and outward as the business landscape changes, Carter said. Leaders should be paying attention to market needs and understanding customers while assessing their autonomous ecosystem.  

Roles all across the enterprise will likely shift during this transformation, Carter said. It will become important for agent management to become a part of workflows, but it will also free up time for teams to focus on the uniquely human parts of business, she said. 

“We can get rid of some of the drudgery,” she said. “What's the specialty stuff that's left over?”

AI will bring deep transformation to nearly three-quarters of business sectors by 2035, putting $4.7 trillion in profits at stake, according to a report from Bain & Co. The global management firm compared profit trends and structural changes from the early era of the internet to understand how the technologies differ in their effect on business operations.

Whereas the internet’s main impact was slashing the cost of distribution, and easing customer access for businesses, AI collapses the cost of production itself, Bain & Co.’s research found.

CIOs who want to understand how their industry will be affected should continue to push their tech transformations, Dunigan O’Keeffe, a partner in Bain & Co.’s strategy practice, told CIO Dive in an email. “They need to place a stake in the ground on the future industry profit pool for their sector, how they will win, and then what that requires from technology,” O’Keeffe said.

AI vendors advertise the technology's ability to accelerate work processes by taking over some human-led tasks and optimizing production in physical settings. 

But while some managers report productivity gains from AI use, the ROI remains hard to measure. On average, automation saved workers about 11 hours a week, Glean’s Work AI Institute found in June, but employees reported spending much of that saved time managing AI outputs.  

The impact AI will have on enterprises is less about the technology’s transformation, and more about business transformation, O’Keeffe said. The technology is driving meaningful change in operational capabilities for companies and the sectors they inhabit. 

AI is set to triple the economic impact of the introduction of the internet in just about a decade, Bain & Co.’s research shows. About 41% of sectors were transformed by the internet between 1995 and 2015, compared to 71% projected to be transformed by AI. 

“The CIO is uniquely positioned to play a leading role as this conversation requires a granular understanding of what the technology makes possible,” O’Keeffe added.

Productivity, innovation, market share and competitive shifts created by AI will provide the most impact on the business sector, Bain & Co.’s research shows. 

Amid this period of transformation, CIOs should plan to build proprietary intelligence, O’Keeffe said — the specific data, technology, and learning systems that will create advantages for their companies. 

Seeking out general productivity gains is good, but could have a smaller effect than chasing tailored, specific business goals that AI can help achieve, O’Keeffe said. Advancements in productivity represent $1.1 trillion in new profits, but innovation and competitive shifts represent $3.5 trillion in profit changes, the report found. 

Tech leaders should benchmark their companies’ speed against the fastest-moving competitors in their sectors, and plan to share frontline learnings that build understanding about the direction of the company’s investments into the technology. 

“The CIO is a catalyst for making the transformation dynamic, keeping the pulse on where the organization is seeing results, the new barriers to attack, and how the frontier of technology is evolving,” O’Keeffe said.

Enterprises are navigating rising AI costs while avoiding vendor lock-in as an increasingly complex infrastructure layer presents additional challenges. 

Vendors are starting to recognize the struggle enterprises face with managing AI infrastructure. Broadcom introduced VMware Private AI Cloud and its software-defined foundation VMware AI Factory. The foundation automates infrastructure deployment capabilities such as hardware provisioning, end-to-end lifecycle management and software stack enablement. 

Integrating hardware and software operations lets enterprises reduce operational complexity while quickly scaling AI workloads, the company said. VMware AI Factory and Private AI Cloud are Broadcom’s “first full-throated response” to helping enterprise customers along their AI transformation journeys, according to Matt Kimball, principal data center analyst at Moor Insights & Strategy. 

With environments growing larger and becoming more challenging to govern, even the best system administrators will get lost, Kimball said in an email to CIO Dive.

“It’s not about capabilities,” Kimball said. “It’s about the complexity of all these new demands tied to GPUs, memory tiering and performance tuning. Especially as we start to see AI emerging as a new workload with such unique requirements.” 

Tools such as the latest platform releases from Broadcom are maturing in a way that will help enterprises address infrastructure complexity, a key friction point for AI adoption. 

“What VMware has delivered is a good opening salvo, but I expect to see VCF — the VMware AI Factory and Private AI Cloud — rapidly evolve into that autonomous AI control plane,” Kimball said.  

Simplifying the infrastructure layer

A concept largely led by Nvidia, AI factories focus on providing advantages of centralized, integrated stacks and giving enterprises all the performance benefits of AI, Brad Shimmin, VP and practice lead of data intelligence, analytics and infrastructure at The Futurum Group, told CIO Dive. At the same time, they enable optionality without introducing complexity and technical debt, he said. 

“It has and will continue to be a multicloud, multivendor, multimodal environment,” Shimmin said. “Every single layer of that AI stack companies are building right now is taking advantage of this ever-moving conveyor belt of innovation that’s coming out of both the frontier model makers and the clouds that are hosting and building on top of these models.” 

VMware AI Factory is a “great move for Broadcom,” which is known for its ability to simplify infrastructure, he added.  

To streamline deployment of the VMware AI Factory, Broadcom is partnering with MetalSoft to reduce time spent on bare-metal provisioning from weeks to minutes, the company said. IT executives will be able to provision physical servers from multiple vendors directly through the VMware Cloud Foundation management console, “unifying the software and hardware lifecycle into a single operational model,” the company said. 

VMware AI Factory also gives enterprises a path to run leading AI models on-premises. Broadcom announced during an Aug. 31 event that frontier AI models from providers including Google, Nvidia and Alibaba Cloud are validated to run on VMware Cloud Foundation, which can be accessed in Private AI Cloud. VCF customers have access to more than 150 open source and commercial models. 

VMware AI Factory enables enterprises to choose both preferred hardware and vetted models, enabling better control over AI tokenomics more broadly, Paul Turner, chief product officer of the VMware Cloud Foundation division at Broadcom, said in an announcement. 

Broadcom’s move to abstract underlying hardware complexity through automated infrastructure capabilities also addresses a lot of skills gap issues, but doesn’t fully solve them, Kimball said. 

“It’s a great first step,” Kimball said. “But a first step nonetheless.”