COMMENTARY: A new report from Omdia found that governance and compliance challenges are still the biggest barriers to AI adoption for most organizations. The research finds that more than half of managed service providers (MSPs) list end-customer governance challenges as the primary blockers to AI adoption, ranking above data and security management (14%), value realization (14%), and technical expertise gaps (13%). [SC Media Perspectives columns are written by a trusted community of SC Media cybersecurity subject matter experts. Read more Perspectives here.]These findings underscore what many in the industry may know already: that successful AI adoption has become an operational problem, rather than a purely technical one. Companies can now make or acquire highly competent AI tools with ease, but we’ continue to see limited and cumbersome adoption of these tools without the right support and operational guardrails. Let’s take a look at the three most important steps that leaders should implement to adopt AI effectively: Agentic AI has been widely deployed in most regions and sectors, but many organizations still lack visibility into the basics of agent activity. Many don’t know what agents are doing with any precision, or even how many agents are deployed within a particular environment. According to one survey, less than half of agents are governed. That same report estimates that there are roughly 3 million ungoverned agents in the UK and U.S. alone. Ungoverned agentic AI poses obvious operational and security concerns. If we don’t know what our agents are doing or how many agents we have, we can’t regulate their behavior. They also create new and frivolous costs, since unmonitored agents are more likely to spend time on needless or redundant tasks, which can make it harder for leaders to show ROI with agentic AI investments. Agentic AI has been largely ungoverned, and that’s a problem. To succeed with agentic AI, organizations need robust governance structures in place before they roll out their tools. Vendors are already creating products to meet this need. 2. Implement integrated governance and compliance, not point solutions. Omdia reports that 49% of MSPs want a complete and integrated platform for governance, compliance, and AI solutions, while almost all (91%) want integrated data backup and recovery. As AI becomes more powerful and closely entwined with adjacent technologies, governance and compliance will become increasingly complex. And that’s why organizations that deliver AI services want integrated platforms instead of disconnected point solutions to implement and manage the technology succinctly. This represents an acceleration of an ongoing trend in enterprise technology. For years, companies have been moving away from point solutions because cost efficiency and operability concerns. As macroeconomic pressures and AI put greater pressure on IT budgets and teams, organizations seem to have less appetite for point products than ever before. Integrated and platform-oriented AI governance and compliance technologies are the most viable path forward for a straightforward but important reason: they do more – and better – work for less. 3. Promote continuous education and training for high-level stakeholders. Organizations need to prioritize ongoing education and ethical training for all leading stakeholders involved in AI adoption and use. Strong education around governance and compliance helps ensure that leaders have a clear understanding of regulatory requirements, responsible AI practices, and the evolving risks associated with Agentic AI. By fostering a culture of awareness and accountability, enterprises can proactively mitigate compliance challenges and build trust in their AI initiatives. It’s also important to implement education programs for AI users at the organization, and research indicates that most organizations have or will soon implement this in some way. According to a recent study, almost all (99%) of organizations that use AI say they have some sort of training program in place. There’s no panacea for the challenges that are stalling AI adoption, but the research shows us that strong governance, compliance, and human training programs are table stakes moving forward. Without robust integrated systems and support, many AI pilots and nascent programs simply won’t work. Scott Sacket, senior vice president, partner strategy, AvePointSC Media Perspectives columns are written by a trusted community of SC Media cybersecurity subject matter experts. Each contribution has a goal of bringing a unique voice to important cybersecurity topics. Content strives to be of the highest quality, objective and non-commercial.
- Establish visibility and accountability for AI agents.
