While there’s a lot of excitement around AI, the reality for many enterprise IT operations (ITOps) teams is increasing complexity, more tools, more data, and more disconnected automation.
Omdia recently surveyed 385 IT and data professionals across North America, examining priorities, adoption patterns, and the real-world impact of AI on ITOps. The findings reveal that enterprises are not just adopting AI—they are rethinking how IT operates and looking toward a new operational model where AI can act with context, coordination, and control.
Omdia Principal Analyst Jim Frey and BMC Helix Chief Product Officer Ryan Manning sat down for a new webinar about what it takes to actually operationalize AI in a way that scales.
Below are highlights drawn from both the research and their conversation, which dives deeper and adds context.
Complexity Is the core problem AI must solve
One of the strongest themes to emerge is that AI adoption is being driven less by novelty and more by necessity. Organizations are grappling with hybrid infrastructures, multi-cloud environments, faster application cycles, and talent constraints. As Frey explained, that’s making investment in AI a survival mechanism. “The number one thing that folks really need help with … is this issue of increasing complexity,” he shared.
Manning added that AI is also a practical way to keep up amid ongoing hiring slowdowns. “99% of CFOs … have tempered their hiring forecast, so there’s not more human labor on its way to help with some of these issues,” he explained.
During the webinar, they explored in more detail how this complexity shows up, and why AI can’t solve it without the right operating model behind it.
From copilots to agents—and autonomous action
While early AI adoption has focused on copilots and human-initiated assistance, the report highlights a clear shift toward agentic AI and conditional autonomy. Many organizations now see autonomous agents as essential for handling the scale and speed they need, as long as guardrails are in place.
As Manning notes, “enterprises need a new operating model … to deliver on the … absurdly high expectations of AI.” While fully autonomous IT operations are still elusive for the majority of survey respondents, Frey shared that the bold early adopters that are doing it “are seeing the most significant improvements in their KPIs.”
The live conversation went deeper into what “conditional autonomy” looks like, and where organizations are drawing the line today.
Measuring value through a “digital workforce ”
Another topic was measuring the impact of AI through productivity metrics and digital labor equivalents. Manning described how BMC Helix operationalizes this. “Once our AI agent has saved up to 2,080 hours of work … we count one digital full-time employee (FTE) created,” he shares. “We’re opening up capacity [and] creating the digital workforce.”
This framing, discussed in detail during the webinar, reflects how CIOs and CFOs increasingly talk about capacity, cost, and accountability, and helps move discussions away from fear-based narratives and toward measurable outcomes. The webinar expanded on this with commentary on financial visibility, cost tracking, and executive alignment.
Cross-domain thinking is where the real payoff lies
Both speakers stressed that the impact of AI deployments can be limited if those deployments are siloed. IT incidents, security, service management, and cloud operations are deeply interconnected, and AI must operate across domains to resolve issues. As Frey says, it’s “where the real payoffs start to show up.”
Another piece of that cross-domain functionality is integration into the familiar tools that everyone is already using. Manning sees a future where collaboration platforms like Teams and Slack “are going to quickly turn into work management platforms.”
During the webinar, they also discussed what breaks down when AI is applied in isolation—and how organizations are starting to reconnect fragmented workflows.
Governance and security remain front and center
Security and governance were explored, as both top concerns and active forces shaping adoption strategies. Frey highlighted how often security teams are now involved early in AI evaluations, while Manning shared how governance requirements affect deployment models and architectural choices.
Rather than slowing progress, this level of scrutiny is increasingly helping organizations define where autonomy is appropriate and where tighter controls are required. During the webinar, audience questions expanded this topic into industry-specific considerations and regulatory realities that influence how far organizations can go today.
Success is measured in myriad ways
According to the survey, enterprises are measuring the key performance indicators (KPIs) of their AI success in terms of degree of autonomy achieved, predictive accuracy, and time savings, as well as traditional ITOps metrics like mean time to detection and restoration (MTTD/MTTR), mean time between failures (MTBF), and return on investment (ROI).
“This tells us [that the] jury’s in. This is working and there’s an incredible amount of value to be had,” explained Manning. “And we think we’re going to get really good as an industry by comparing ourselves to your peers [and] how [they] are performing … they’re going to get pretty jealous of some folks that have decided to lean in for the technology.”
The conversation also explored how early adopters are pacing their investments and where they are seeing the fastest returns.
Parting thoughts
The discussion ultimately reinforced that AI does not reduce complexity on its own. Real progress happens when organizations pair AI with new operating models, cross domain thinking, and clear governance.
BMC Helix. (2026, April 24). From AI experimentation to operational reality: Turning complexity into control. BMC Helix Blogs. https://blogs.helixops.ai/webinar-agentic-ai-complexity-control/