The growing adoption of artificial intelligence by people and companies has definitively diverted the debate about its use. Companies are no longer wondering if they should incorporate AI, but rather how to do it without losing control over the most sensitive information they handle and the data they use in their processes. The AI at Work study (BCG, June 2026) states that, internationally, 74% of frontline employees use AI in their work activities several times a week.
This figure increases considerably when the level of responsibility increases: 88% of managers and 93% of leaders use AI regularly. These profiles handle, on a recurring basis, sensitive and confidential information, which gives weight to secure and trustworthy processing and storage needs. In this context, guaranteeing the protection of the data handled by organizations becomes a priority.
Controlling data, the challenge for adoption
Although there is a growing willingness to incorporate AI in companies, trust is the real brake on taking the leap. According to the National Observatory of Technology and Society (ONTSI)(2), the most cited reason for not adopting AI is lack of knowledge about how to protect it. A well-founded concern, especially when AI stops answering questions and starts acting on its own.
An autonomous agent can operate with the employee’s own credentials, communicate with external services without validation or mix information from different processes. These activities give agents access to vulnerable and sensitive data in the face of cyberattacks and leaks. Therefore, controlling what data AI systems can query, process or share is essential to maintaining the security of business information.
For Juan Antonio Monge, B2B country product manager at ASUS, “The adoption of artificial intelligence is advancing very quickly and, with it, also the quantity and sensitivity of the information that we put at your disposal. When we use AI to analyze internal documents, work with customer data or manage confidential information, it is no longer enough to ask ourselves what it can do: we must also know where that data is processed and how we maintain control over it.”
This same concern about controlling where the data lives is repeated outside of Spain. According to the report State of AI in the Enterprise: The Untapped Edge (Deloitte, January 2026)(3), 83% of international companies already consider the residency of their data a key factor in their strategic planning.
Local AI as a response to the challenge of data control
Given this scenario, ASUS focuses on offering companies greater control over how and where their information is processed. One of the keys is local AI, which allows certain loads and artificial intelligence models to be executed directly on the device, reducing the need to send information to external servers for processing and, thereby, limiting data exposure.
“The next evolution of AI in the company is not only about doing more things, but about deciding better how we do them. Processing certain AI tasks on the device itself allows sensitive information to remain on the computer and gives organizations greater control over their data,” says Juan Antonio. “Therefore, the ability to run AI locally is no longer just a question of performance and is also becoming a criterion when choosing technology for the professional environment.”
This vision is materialized in computers like the ASUS ExpertBook Ultra, prepared to execute AI tasks directly on the device. To do this, it incorporates an Intel® AI Boost NPU of up to 50 TOPS, exceeding the 40 TOPS required by Microsoft for Copilot+ PCs, and reaches up to 180 TOPS of combined power between NPU, CPU and GPU.
A power that translates into specific day-to-day tasks for a consultant or manager, such as summarizing a meeting by identifying each interlocutor, working with confidential documents or analyzing information and drawing conclusions, reducing the need to send this data to external services.
ASUS focuses on offering companies greater control over how and where their information is processed
But information control does not depend solely on where the AI is executed. For this reason, local processing is complemented by ASUS ExpertGuardian, the ExpertBook Ultra security system, which incorporates different layers of protection to reinforce the security of the data and the device itself. These include the TPM 2.0 chip, fingerprint access with FIDO2, the Microsoft Pluton coprocessor and Opal encryption on the SSD, which helps protect stored information even if the device is lost or stolen.
“Companies are going to continue incorporating new artificial intelligence capabilities and the challenge will be to do so with the same guarantees that they require from the rest of their critical processes. Innovation and security have to advance at the same pace so that AI can be integrated naturally and sustainably into the day-to-day life of organizations,” concludes Monge.
