AI is already being used in all companies, whether it has been formally approved or not. Workers are using AI with or without IT involvement, fueling the rise of “shadow AI” in companies, creating gaps in governance and control.
This situation is no longer a simple IT challenge. For information security managers, this change increases the attack surface between devices, endpoints and data flows, introducing unmanaged risks and increasing the likelihood that sensitive company data will be accessed or exposed without proper oversight. This situation reveals a growing gap in AI execution: use is accelerating, but control is not keeping pace.
After surveying more than 6,000 workers around the world, the latest report in Lenovo’s Work Reborn series, Leading Your Workforce to Triumph with AI, finds that more than 70% of employees use AI weekly and up to a third operate without IT supervision. At the same time, 80% anticipate their reliance on AI will increase over the next year.
«AI adoption is no longer the main challenge. “Now it’s about execution,” said Rakshit Ghura, vice president and general manager of digital workspace solutions at Lenovo. «Usage is increasing faster than companies can control or ensure. Without that control, AI introduces as many risks and costs as it does opportunities.
Uncontrolled AI is already impacting business performance
When AI use scales without visibility or governance, the impact is no longer theoretical. Today, it is already affecting costs, level of security and ability to extend AI throughout the business.
Companies are experiencing:
- Delays in return on investment as AI initiatives remain fragmented across teams.
- Duplicate expenses, when using multiple tools to solve the same problems across silos.
- Increased attack surface, generated by the use of unapproved tools to access business data.
- Lack of visibility, which makes it difficult to scale those systems that work correctly.
At the same time, AI adoption among workforces is very uneven. While some workers operate in secure, optimized environments, others turn to any tool they can access to stay productive. This results in workforces operating at two different speeds, which slows down decision-making, duplicates efforts, and makes it difficult to adopt AI consistently across the enterprise.
Uncontrolled AI is expanding the attack surface faster than security can accommodate
As the use of AI accelerates, the risks also escalate in parallel. Although 61% of IT managers report an increase in cybersecurity threats associated with AI, only 31% are confident in their ability to manage these risks. On the other hand, 43% of workers express concern about data exposure or AI-driven attacks.
Without a clear governance framework, AI is quietly expanding the enterprise attack surface, increasing the likelihood of breaches, compliance failures, and operational disruptions.
The problem: fragmented AI management
Most companies are trying to manage AI across disconnected layers. Devices are deployed and managed in one way. The infrastructure is implemented in a different way. Security is often added later, as an additional layer. All of this fragmentation is what is causing the AI execution gap.
Adding more tools and policies does not solve the problem. All it does is increase complexity, create gaps between endpoints and infrastructure, and make it difficult to apply consistent control across the entire environment.
Lenovo’s approach: control AI at the device level and apply security as a service.
Lenovo has taken a fundamentally different approach. Control is established at the first point where AI connects with the company: the devices.
From there, Lenovo links device deployment, lifecycle management, infrastructure and security into a single operating model governed and implemented through its device-as-a-service platform for security, TruScale Device as a Service for Security.
It is not simply a combination of technologies, but a fully managed service that combines:
- Enterprise-grade devices, guaranteed from day one.
- Integrated device and firmware protection using Lenovo ThinkShield
- Advanced endpoint security from leading partners.
- 24/7 managed security services, including monitoring, detection and response.
Most entities are forced to put together and operate these types of systems on their own, using multiple different vendors. Instead, Lenovo offers it as a single, comprehensive, managed service, thereby reducing complexity and closing gaps across the entire environment.
Because security is built in from the moment of deployment and actively managed over time, companies can:
- Reduce risks through constant monitoring and proactive response.
- Eliminate gaps between device security and operational security.
- Simplify supplier management and reduce the total acquisition cost.
- Free up internal IT and security teams to focus on higher-value initiatives.
This approach is what sets Lenovo apart. Instead of managing devices, infrastructure and security separately, Lenovo applies a continuous control model across the entire environment, something other vendors cannot offer in an integrated way. Deployed through a flexible As-Service model, this approach aligns AI investment with actual demand, reducing upfront costs, avoiding duplication of expenses, and scaling security devices and services as AI adoption continues to evolve.
Lenovo’s approach is to control AI at the device level and apply security as a service
Lenovo’s approach to cybersecurity and device protection has been recognized by external sources, such as the recent Fortress Cybersecurity awards.
How to close the AI execution gap and start realizing ROI faster
More than 70% of employees recognize the potential of AI to generate gains in productivity, speed and quality. However, realizing this value depends on execution.
Those companies that manage to close the AI execution gap will be able to move from fragmented experimentation to measurable results more quickly. This will reduce expenses, limit risks, and create a clear path to scale AI across the business.
When devices, infrastructure and services operate under a unified and managed model, AI goes from an unknown vulnerability to a controlled and scalable advantage.
