This voice is automatically generated. Please let us know if you have any feedback.
As organizations rush to adopt AI, technology has outpaced monitoring capabilities. TrustedTech’s two-part report says many are now grappling with fundamental governance issues all the way to the top, with organizational leaders becoming prime suspects in the shadow AI hunt.
Microsoft’s cloud solutions provider analyzed responses from 2,001 UK and US employees surveyed in the March census. More than three-quarters of decision makers said they were confident in effectively leveraging AI in their roles. Only 43% of employees at decision-maker level and below said the same, according to the second part of the report published Tuesday.
Meanwhile, 36% of respondents said they learned AI primarily on their own, and only 23% said it was through training provided by their employer. Among U.S. respondents, 46% said their organization lacks adequate training on safe and secure use of AI. Across the entire sample, 41% said there was a lack of clear guidance on AI in the workplace, and 48% said employers should take primary responsibility for upskilling.
TrustedTech says the findings point to a readiness issue, not a worker issue.
“Organizations have prioritized the adoption of AI tools over preparing the people who will use them,” said Julian Hammoud, founder of TrustedTech. “The workers most at risk of being laid off receive the least investment, and this is no coincidence. The trust gap our data reveals is not a talent problem. It is a leadership failure, and there are solutions.”
The study helps explain some eyebrow-raising data points revealed in the first part of the report, published in May. Executives are disproportionately responsible for the proliferation of shadow AI. Nearly two-thirds of decision makers admit to using unapproved AI tools, and the rate is more than double among lower-level employees.
“The initial announcement from our research shows that shadow AI is not primarily being driven by junior employees experimenting with ChatGPT, for example,” Andy Nolan, vice president of technology at TrustedTech, told Channel Dive. “It is driven by senior leadership at the highest levels, who are making strategic decisions while often operating outside of established governance.”
The latest edition of the report sheds further light on the issue. Nolan said that while companies are proactive about the tools, they are not addressing the training, oversight and cultural foundation needed to make sanctioned AI adoption viable across the business.
Governance first
A standard partner handbook for platform deployment, policy publication, and end-user training is not enough for AI.
“Our research shows that executives are often not only the earliest adopters of AI, but also the ones most frequently using unapproved AI tools,” Nolan said. “It’s not because they’re intentionally ignoring security. It’s because they’re trying to move faster than their governance processes will allow.”
He added that in many organizations, leadership sets cultural standards for AI adoption before formal policies exist. For partners, that means implementing top-down oversight.
“Executive teams need to understand both opportunities and risks, because their actions set the tone for the entire organization,” Nolan said. “If executives are not included in governance discussions from day one, they are often the first users to unintentionally circumvent governance.”
Organizations need to establish governance before or during implementation.
“When Microsoft Copilot is deployed before an organization has established data access reviews, acceptable use policies, and governance guardrails, shadow AI behavior often continues at all levels even after approved AI tools are available,” Nolan said.
Employees should be trained based on their role, not their title.
“Front-line and mid-level employees, those whose roles are most susceptible to AI disruption, have the least organizational support,” Nolan said. “Different departments, different workflows, and different risk profiles require different guidance and real-world use cases.”
Beyond the license
Workers are not rejecting AI. TrustedTech has found that we use it because it works. Seven in 10 respondents said AI had a positive impact on their performance, and more than half said they saved at least three hours per week using AI tools.
In this context, restrictions on the use of AI become an issue. More than a quarter of respondents admitted that they would continue to use AI even if it were banned in the workplace. More than a third of decision makers expressed a similar opinion.
Channel partners with expertise in AI deployment face significant opportunities that go far beyond deployment and licensing. Customers want governance frameworks, role-based deployment planning, security controls, acceptance policies, executive advisory services, employee training, and continuous measurement.
“Purchasing an AI license is not the same as changing the way your organization works,” Nolan says.
Additionally, partners need to measure whether employees are actually using approved tools confidently, securely, and in a way that produces business results.
“Customers aren’t just asking, ‘Can I implement this technology?'” Nolan says. “They’re asking, ‘How can we make AI successful across the organization?’ This is a deeper, longer-term, more meaningful path to engagement.”
The most powerful partners do more than just sell the most AI seats. It also enables increased productivity across the organization while mitigating the risk of shadow AI.
Even as organizations seek to balance innovation and governance, employees want to be more productive and executives are looking to move faster. These priorities are not inherently contradictory, but they must be aligned from the beginning of an AI launch.
“Organizations don’t need more restrictions when it comes to AI,” Nolan said. “They need a trusted pathway that facilitates safer AI deployments than unapproved alternatives.”
