Enterprise leaders are confronting growing concerns about AI security, data exposure and agentic risk, but those threats do not appear to be stopping adoption.
New research and conversations with managed service providers instead point to a shift in how businesses are deploying AI: moving forward with more emphasis on governance, security and oversight. Deloitte found 93% of surveyed North American CFOs are already using AI across multiple key functions, even as 59% said balancing rapid deployment with potential risks is their biggest AI governance challenge.
Peter Rodenhauser, president and CRO of Corsica Technologies, said the midmarket customers his company serves are following a similar pattern.
“I wouldn’t say that we’re seeing a slowdown, per se,” Rodenhauser told Channel Insider. Instead, customers are piloting AI implementations, measuring their value and determining where to expand next.
For MSPs, that tension is creating an opportunity to move beyond securing technology and into deeper conversations about governance, business processes, and how customers deploy AI safely at scale.
Enterprise AI adoption continues despite security risks
Deloitte’s Q2 2026 CFO Signals survey illustrates the tension facing enterprise leaders.
Cybersecurity ranked as the second-most-cited external AI concern, identified by 41% of respondents.
For Rodenhauser, that tension makes security and governance prerequisites for expanding AI rather than arguments against adoption.
“You’re managing risk at the end of the day when you’re deploying this,” he said. “How much risk do you want to take on or have an appetite for? That’s going to determine what policies you need to put in place.”
Corsica tries to establish that foundation early.
“We really did try and lead with governance and policy as a conversation,” Rodenhauser said, noting that policies can then evolve as organizations expand their use of AI.
READ MORE: Our guide to AI governance services explores how MSPs can operationalize and scale their offering.
AI governance is advancing, but security gaps remain
Research from IANS Research and Artico Search suggests many security leaders are taking that approach.
In a survey of 113 CISOs conducted in April and May, 66% said their organizations have a dedicated AI policy. But policy development hasn’t eliminated the technical risks: 74% of enterprise AI environments already pull data from external sources, while only 29% of organizations have conducted adversarial testing of their AI systems.
“AI is not going to fix problems for you that humans could not have,” Promevo CEO Karthik Kripapuri told Channel Insider recently. “What it’s going to do is shine a massive light on…inefficiencies and lack of proper process documentation and governance models.”
Those gaps could become more consequential as agentic AI adoption accelerates and businesses move from employee productivity tools toward AI agents capable of interacting with enterprise systems and data.
“I think it’s going to be in the center,” Rodenhauser said of security. “I think it has to be in the center because, as there are more agents being deployed… there are unintended consequences, and that’s really what we’re trying to protect ourselves against.”
PwC’s Global Digital Trust insights survey suggests that those leaders are also putting their money where their mouths are: 84% expect their cybersecurity spending to increase over the next 12 months.
MSPs can move further into the business
For channel partners, the shift creates an opportunity that extends beyond securing AI tools.
Rodenhauser said Corsica’s customer conversations have already progressed from basic questions about where to start toward increasing adoption, generating more value and developing agents tailored to specific business processes.
That changes the MSP relationship.
“I believe MSPs are starting to go further up the value chain within an organization, because we are inherently having conversations more around business processes,” Rodenhauser said.
He pointed to discussions with HR departments about using agents to improve employee onboarding as one example—conversations traditional infrastructure-focused MSPs might rarely have entered.
The emerging opportunity, then, isn’t simply helping customers choose between AI innovation and security. It’s helping them achieve both.
“Continuing to balance that protection with governance, protecting the business, with advancing the business through this new technology, is just going to be key,” Rodenhauser said.