A recent report from SailPoint reveals that while 96% of IT professionals perceive artificial intelligence agents as a security risk, a staggering 98% of them still plan to expand the use of these agents within their organizations over the next year. The study surveyed over 350 IT professionals and found that 84% of respondents currently utilize AI agents, but only 44% have established policies to govern their behavior. These autonomous systems, which can operate independently without human oversight, are becoming prevalent in the wake of the generative AI boom initiated by the release of ChatGPT in late 2022. Chandra Gnanasambandam, SailPoint’s Executive Vice President of Product and Chief Technology Officer, highlighted the need for stringent governance, stating that while 92% of respondents recognize the importance of adequate oversight, 80% have reported agents acting in unexpected and potentially harmful ways. This highlights the urgent necessity for organizations to implement strict security protocols akin to those applied to human employees.
A recent survey by Gartner reveals a growing dissatisfaction with cloud computing among organizations, attributing much of this discontent to unrealistic expectations and improper implementation. The consultancy estimates that up to 25% of organizations may face significant disappointment in the coming years due to unexpected costs and management complexities. Gartner’s advisory director, Joe Rogus, emphasizes that many organizations lack a coherent cloud strategy, which leads to issues like vendor lock-in and rising expenses. For example, the company 37Signals recently faced a staggering $3.2 million bill for cloud services, prompting it to migrate back to on-premises infrastructure. Furthermore, a study found that typically only 13% of provisioned processing power and 20% of memory is utilized in Kubernetes deployments, highlighting inefficiencies in resource management. As organizations increasingly adopt multi-cloud strategies, Gartner warns that more than half may not achieve their expected outcomes due to technical barriers.
UK Information Technology leaders are increasingly wary of the influence of major United States cloud providers, with over 60% of respondents in a recent survey by Civo advocating for a reduction in the government’s reliance on these services. The survey reveals that 61% of IT leaders now prioritize data sovereignty, and 45% are considering repatriating workloads from the public cloud due to ongoing concerns about data protection. Trust in US-based technology firms has significantly declined, driven by political tensions and regulatory compliance issues. Only 36.6% of survey participants express confidence in these companies to manage their data responsibly. As data sovereignty gains traction, hyperscale providers like Microsoft, Google, and Amazon Web Services have begun to adapt by offering sovereign cloud services tailored to local regulations. This shift indicates a critical evolution in how organizations view data management and security in a rapidly changing landscape.
Why do we care?
IT decision-makers are simultaneously accelerating the adoption of autonomous technologies while losing trust in the infrastructure that supports them. This contradiction reveals a widening governance gap — and opportunity — for IT services firms.
One could argue that doubling down on AI agents while repatriating infrastructure is incoherent. Why build cutting-edge autonomous systems if you’re regressing to legacy infrastructure? The deeper story may be that organizations are overcorrecting in both directions — overspending on innovation without planning, and over-retreating when complexity overwhelms.
Governance around AI agents — not just usage, but autonomy, data access, and lifecycle control — is an emerging managed service category. This isn’t traditional endpoint management; it’s identity, access, and intent control for non-human actors. Services firms that move quickly to productize policy implementation, monitoring, and agent lifecycle management can build long-term strategic relationships.