AI-Edge-Deployment-Resilience

Network resilience concerns delay AI, edge projects at 72% of organizations

AI News

Organizations with stronger network recovery capabilities reported $6 million lower average outage costs and fewer deployment delays, according to a survey of 514 CIOs and CISOs

Network resilience concerns have forced nearly three-quarters of organizations to delay, scale back or pause AI and edge infrastructure projects over the past 12 months, highlighting the growing importance of reliable network access and recovery capabilities as enterprises expand distributed computing environments.

According to new research from Opengear, 72% of organizations surveyed had delayed, reduced the scope of or paused an AI or edge project because of concerns about network resilience. The findings suggest that weaknesses in network infrastructure are becoming a constraint on enterprise AI and edge deployments, even as organizations increase investment in these technologies.

The research also found a substantial difference in financial exposure between organizations with stronger network resilience capabilities and those with less developed arrangements. Organizations classified as having embedded network resilience reported estimated average outage costs of $12.4 million over the past 12 months, compared with $18.8 million among organizations with less developed capabilities.

That represents a $6.4 million difference in estimated outage costs.

Opengear, a Digi International company, surveyed 514 CIOs and CISOs across the US, UK, Germany, India and Australia in July and August 2026. The research examined how organizations maintain access to and recover critical infrastructure when their primary network becomes unavailable.

For the study, Opengear divided respondents into two groups: organizations with embedded network resilience and those with more reactive approaches. The classification was based on three measures: the ability to remotely access and recover critical infrastructure when the primary network is unavailable, how frequently recovery procedures are tested without the primary network, and how heavily organizations depend on sending engineers onsite to resolve infrastructure problems.

The differences between the two groups extended beyond outage costs to project deployment, routine infrastructure management and cybersecurity incident response.

Resilience affects AI deployment timelines

Organizations with embedded network resilience were roughly half as likely to experience deployment delays of four days or more. Only 21% of respondents in this group reported such delays, compared with 39% among organizations with less developed resilience capabilities.

The research also found that organizations with stronger resilience arrangements were nearly twice as likely to have secure remote access and recovery capabilities in place before deploying AI or edge infrastructure.

Deployment readiness also differed between the groups. Nearly all organizations with embedded network resilience had already deployed AI or edge infrastructure, compared with approximately four in five organizations in the reactive group.

The findings indicate that network resilience is not simply an infrastructure consideration that becomes relevant when systems fail. It can also influence how quickly organizations can introduce new technologies and expand distributed environments.

Routine infrastructure management revealed another significant gap. Nearly half (49%) of respondents from organizations with embedded resilience said they were very confident their teams could maintain access, visibility and control during routine infrastructure changes. Among organizations with less developed capabilities, the figure was 28%.

This suggests that the benefits of network resilience extend to day-to-day operations, where infrastructure changes can create availability and access challenges even without a major outage.

Stronger recovery capabilities improve cyber incident readiness

The differences were also evident in organizations’ ability to respond to cybersecurity incidents.

During a cyber incident, 93% of organizations with embedded network resilience said they were confident they could isolate affected systems while retaining access to unaffected infrastructure. The corresponding figure among organizations with less developed resilience capabilities was 74%.

Maintaining independent access to infrastructure during an incident can be important when primary networks or management systems become unavailable or are compromised. Without alternative access and recovery mechanisms, IT teams may struggle to isolate affected systems, investigate incidents and restore services.

Opengear President Patrick Quirk said organizations have traditionally viewed resilience primarily as a response to infrastructure failures, but the research points to its broader operational value.

“The industry has traditionally treated resilience as a response to failure. This research shows it is also a source of performance,” Quirk said in a statement.

As enterprise networks extend across data centers, cloud environments, branch offices and edge locations, maintaining secure access and control across these distributed environments is becoming increasingly important, he said.

AI and edge expansion drive resilience investments

The survey also found that organizations with stronger resilience capabilities were more likely to view network resilience as a strategic requirement for future infrastructure expansion.

Some 84% of respondents from the embedded resilience group agreed that network resilience was becoming more important as AI and edge infrastructure expanded. Among organizations with less developed capabilities, 69% shared that view.

Supporting the growth of AI, edge and distributed infrastructure was also the leading investment priority for the embedded group over the next 12 months.

Quirk said organizations need to establish the ability to access, monitor and manage infrastructure before deploying new technologies, rather than treating these capabilities as something to address after an outage.

“AI and edge are becoming more prominent in the business, and every organization has the same question: can I see it, reach it and manage it?” Quirk said.

Organizations that build resilience into their architecture from the beginning have access, visibility and control mechanisms in place before deployment, giving them an advantage as their infrastructure expands, he added.

The findings come as enterprises increasingly distribute computing and AI workloads across cloud, data center and edge environments. This expansion increases the number of infrastructure locations that IT teams must monitor and manage, making the ability to maintain secure access during network disruptions an important consideration for enterprise infrastructure planning.

Leave a Reply

Your email address will not be published. Required fields are marked *