As businesses race to adopt artificial intelligence and advanced automation, a new report from Broadcom suggests that many companies are building these systems on shaky ground. The survey, “The Orchestration Accountability Gap: The Cost of Poor Governance,” reveals that a lack of oversight, fragmented tools, and unclear accountability are creating big risks for IT departments today—risks that could easily grow into a full-blown AI governance crisis tomorrow.

Broadcom surveyed over 500 enterprise professionals to understand how they are managing their automation tools, and the answers were in some ways contradictory.  

When confidence doesn’t match results

Aline Gerew, Head of Automation, Agile Operations Division at Broadcom, found it surprising that while 84% of IT teams responding say they are confident they’re meeting the business needs, but then 57% they are missing their SLAs on a regular basis. “It’s interersting that they can’t pull it together,” she told ITOps Times, “so the confidence is very disconnected from reality. You can’t have both.”

These results paint a picture of operational complexity that’s becoming too hard to handle. Among those issues are tool sprawl and a lack of end-t0-end visibility into operations so organizations struggle to get a handle on how much they are actually spending on tools. While companies are buying more software to handle their orchestration needs, this has had the unintended effect of making it harder to see the big picture. According to the survey, 97% of companies use multiple orchestration tools, and 60% are juggling four or more.

Instead of gaining transparency, IT teams are losing it. A massive 80% of those surveyed said that their current collection of tools actually makes it harder to get clear, end-to-end visibility into their operations. This “visibility crisis” is a major concern, as 92% of respondents agreed that a unified solution capable of seeing across all these different tools would help reduce operational issues. And, even when you get visibility into problems, some 52% of respondents say their root cause analysis is manual, so when something does happen, it’s taking at least one hour to even figure out what happened, never mind remediating the problem, Gerew said. And, if all of these tools are disconnected, and you’re doing this with each tool, you have to have the expertise to know these different orchestration tools, how they’re used, and to be able to look at what the problem is, and then fix it.” And the longer the problems remains, the risk to the business increase. 

“The problem with the tool sprawl,”Gerew said, “is you don’t have this control plane that you can go to to see at any given point, where are you at? Are you going to miss your SLA?,  You miss your SLA, then you’ve got to go figure out what happened upstream, and then what were the downstream effects, and but it’s always reactive, and you need to really change that model to be more predictive. And the only way to do that is through a central view that is giving you integration into all of these orchestration tools, and then being being able to predict that we’re running late, or this orchestration didn’t work.” This leads to making decisions on incorrect data, or data that isn’t even there yet, she said.

The Rising Risk of Compliance Issues

Perhaps the most pressing concern for the near future is the issue of auditability. The study found that the “accountability gap” is already causing problems, with 89% of companies reporting that they have faced audit issues related to their orchestration processes.

Compliance is difficult without proper documentation. Only 54% of companies say they have the necessary audit trails to track their complex automation workflows. As regulations around AI and data usage tighten, organizations without these traceable, consistent records will face increased risks.

Broadcom’s report concludes that these issues are not just background noise; they are business risks that will likely widen as AI becomes more central to daily operations. The core message for IT leaders is clear: by strengthening governance, visibility, and auditability now, companies can create a more sustainable and responsible path forward for their AI initiatives.

“Too many organizations are building their automation and AI strategies on top of fragmented tools and incomplete visibility,” Gerew said. “This research shows that governance needs to catch up with innovation. The good news is that with the right controls, auditability, and shared accountability in place, companies can move faster with AI and automation and do it with confidence.”