New Cisco Research: CEOs want faster AI investment

AI is exceeding most CEOs’ expectations and their optimism is growing. Cisco’s latest study shows that 65% now worry they’re underinvesting in the technology — up from 53% last year — with more than two thirds (69%) of those surveyed now seeing AI adoption as mandatory for modern business. But significant challenges remain for companies struggling to build the infrastructure, data foundations, and security controls to make AI work safely, reliably and at scale. 

This is the second consecutive year Cisco has surveyed 2,500 CEOs across 23 countries to understand how they view AI, what they expect from it, and what’s getting in the way of scaling.

As AI enters the workforce, most CEOs expect humans to work side by side with AI

Deploying AI agents to work alongside employees is among the top three priorities for CEOs in 2026. By 2030, almost every CEO expects AI to play a role in business, but they don’t expect AI to fly solo. 72% envision a future where AI only supports or executes under human direction, judgment, or governance. Their reasons are both practical and relatable: keep AI systems secure, ensure productivity in human-AI teams, and navigate the ethics of AI decision-making effectively. One CEO put it plainly: “Move fast with AI, but anchor decisions in human values — always.”

CEOs have done the learning, now it’s time to build 
The knowledge gap is closing fast. In 12 months, the share of CEOs who said their understanding of AI held them back in the boardroom fell to 47% from 74%, with those who said it blocked informed decision-making dropping to 49% from 74%. While knowledge and confidence are rising, so is their awareness of what it takes to get this right.  

The execution reality check 

While CEO intent is growing, execution is stalling because of three major constraints: 

• Infrastructure isn’t ready: 53% of CEOs fear infrastructure limits will hold them back, and upgrading infrastructure to handle AI workloads ranks as the top priority for CEOs in 2026, closely followed by upskilling teams for AI readiness.

• Trust is still a concern: Justas CEOs are pushing ahead to make AI agents part of their workforce – deploying agents is among their top three 2026 priorities – security and control of these autonomous systems has become their number one concern.
• Data is still too fragmented to power AI properly: Data quality, accessibility, and centralisation issues blocking AI progress represent the single greatest barrier, cited by more CEOs (34%) than any other challenge.

These CEO concerns are validated by data on the ground. Cisco’s 2026 AI Readiness Index, based on 8,000+ IT leaders globally, shows that organisations globally are facing the same three barriers: less than a quarter (22%) rate their network as optimal for AI workloads, just 31% feel equipped to secure and control AI agents, and only 19% have fully centralised, AI-accessible data. Those who have solved all three are pulling ahead, while the rest risk missing out.