AI left the cloud and now lives in your pocket
Artificial intelligence is no longer that distant resource you accessed in a browser or used for quick questions with a chatbot. What MWC 2025 made abundantly clear is that this technology is moving into a much more intimate territory, deeply woven into our everyday lives. It now lives inside smart devices, integrates with carrier network infrastructure, and holds a central place in the strategy of companies across all sizes and industries. Glasses with contextual awareness, wearables that interpret the surrounding environment, and assistants that make real-time decisions without anyone needing to open an app were some of the demos that turned heads at the event. The overall feeling is that we are standing at the edge of a transition that, while it might seem subtle at first glance, completely changes the way we interact with technology.
When AI stops being a tool you access on demand and becomes an ambient partner — something that accompanies you in the field, at the office, on the production line, or even during a walk — the impact goes way beyond productivity gains. New challenges around enterprise security emerge, new forms of value creation take shape, and naturally, trends surface that no company can afford to ignore. This is exactly the scenario that deserves attention, from the risks of so-called agentic AI to deeper reflections on corporate purpose in the age of intelligent automation. 🚀
NIST steps up to standardize agentic AI security
Before diving into the technical details about the risks that agentic AI presents, it is worth putting into context an important institutional move mentioned in recent discussions. NIST — the National Institute of Standards and Technology — launched a new initiative specifically aimed at creating security standards that cover autonomous artificial intelligence systems. This is not a generic action about AI in general. It is a targeted effort to deal with agents that operate beyond direct human approval, meaning systems that decide, execute, and interact with other systems on their own within corporate environments.
The relevance of this initiative is hard to overstate. Until now, most AI regulation focused on issues like algorithmic bias, transparency, and data privacy. All of those points remain important, of course. But the arrival of agentic AI in the business environment introduces an entirely new category of concerns. When an intelligent agent can, for example, negotiate contracts with suppliers, redirect logistics flows, or adjust production parameters without a human approving each step, the consequences of a failure or a malicious exploit become far more significant. The fact that NIST is addressing this proactively signals that the global tech ecosystem recognizes the urgency of the matter.
For companies operating in international markets or using AI platforms from global vendors, keeping a close eye on this standardization effort is a strategic move. Even if local legislation does not yet require compliance with NIST standards, adopting best practices aligned with these frameworks reduces operational risk and strengthens the overall enterprise security posture. On top of that, when these standards solidify — and that tends to happen faster than many expect — organizations that are already aligned will have a clear competitive edge when it comes to closing partnerships, attracting investment, and demonstrating technological maturity.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
Agentic AI and the new risks for enterprise security
One of the topics that generated the most discussion in the recent corporate AI landscape was the concept of agentic AI — artificial intelligence systems capable of making decisions and executing actions autonomously, without relying on explicit user commands. Unlike traditional models that answer questions or follow well-defined scripts, these agents can interpret complex contexts, plan sequences of actions, and even interact with other systems to accomplish objectives. The potential is enormous, but the risk is proportional. When an AI starts operating with autonomy inside an organization, the attack surfaces multiply. A misconfigured agent can make poor financial decisions, share sensitive data with unauthorized systems, or create vulnerabilities that simply did not exist before in the corporate environment.
For enterprise security professionals, this scenario represents a paradigm shift. It is no longer enough to protect servers, endpoints, and access credentials. Now it is necessary to monitor and govern digital entities that make decisions on their own. Companies that have already started adopting governance frameworks for agentic AI are getting ahead of the curve, creating oversight layers that define clear boundaries for these agents. The point is not to prevent the adoption of this technology, because it is going to happen regardless, but to ensure that control, auditing, and rollback mechanisms are functioning in real time. Organizations that treat security as a native component of AI implementation, rather than a layer added after the fact, will be in a much more comfortable position in the years ahead.
Another relevant point is the need to educate internal teams about these new risks. Many professionals still see artificial intelligence as something controlled by a specific technology department, when in reality it is spreading across every area — from marketing to legal, from finance to logistics. Every employee who interacts with an intelligent agent needs to understand at a basic level how it works, what its limitations are, and in which situations human intervention is essential. Ongoing training is no longer a nice-to-have — it has become a fundamental necessity for any company that wants to adopt AI responsibly and securely.
Corporate purpose redefined in the age of artificial intelligence
A highlight that deserves special attention came from the Harvard Business Review Strategy Summit, where business leaders debated how the purpose of organizations needs to be rethought in light of AI advancement. The central takeaway was clear: companies that continue defining their purpose exclusively around profit and operational efficiency are losing relevance. In the age of intelligent automation, the differentiator lies in articulating a mission that goes beyond financial results and places human value creation at the center of the strategy.
This does not mean abandoning the pursuit of results. It means understanding that artificial intelligence works best when it is designed to augment human capabilities, not simply replace them. Organizations that embrace this mindset are able to attract more qualified talent, build more relevant products, and maintain stakeholder trust over the long term. When automation is guided by a clear purpose — whether it is improving community health, making education more accessible, or reducing the environmental impact of supply chains — it gains legitimacy and scale in a way that purely cost-driven initiatives can never achieve.
In practice, this translates into concrete decisions about how AI is implemented within companies. Product teams that start by asking what human problem are we solving before deciding which machine learning model to use tend to create solutions that are more robust, more ethical, and more commercially sustainable. This alignment between purpose and technology is not idealism — it is a business strategy that delivers measurable results.
Smart devices as value creation platforms
If smart devices were once seen as interesting gadgets for end consumers, they are now establishing themselves as true value creation platforms for businesses. MWC 2025 showcased cases where glasses equipped with contextual AI help maintenance technicians identify problems in industrial equipment just by looking at them. Wearables connected to 5G networks monitor the health of workers in hazardous environments and send automatic alerts when they detect abnormal patterns. Sensors embedded in fleet vehicles analyze road conditions, driver behavior, and the mechanical state of the vehicle simultaneously, all processed at the network edge without needing to send data to the cloud. The common thread across all of these examples is clear — intelligence is increasingly distributed, closer to where things actually happen.
This decentralization of AI brings a direct benefit that often goes unnoticed: the dramatic reduction in latency. When processing happens on the device itself or at the network edge, responses are practically instantaneous. In scenarios like remotely assisted surgeries, heavy machinery operation, or urban traffic management, that difference of a few milliseconds can literally be a matter of life or death. For companies, this means that value creation no longer depends solely on having great algorithms running in distant data centers. It depends on having the right infrastructure in the right place, with smart devices capable of processing, deciding, and acting locally. It is an architectural shift that redefines how we think about enterprise computing.
From a strategic perspective, organizations that can see their connected devices not as operational costs but as intelligence-generating assets will unlock opportunities that the competition simply will not see. Every sensor, every wearable, every smart camera deployed in operations is a data collection point that, when combined with well-trained AI models, can reveal patterns invisible to the human eye. A factory that monitors vibrations in its machines with smart devices can predict failures weeks before they happen. A retail chain that analyzes customer flow in real time can dynamically adjust store layouts and pricing strategies. The technology is already available — what separates those who create value from those who fall behind is the ability to connect these dots in a smart and intentional way.
Trends that will define the next technology cycle
Among the most notable trends that emerged from recent discussions on enterprise AI, one deserves special attention: the convergence of AI, advanced connectivity, and corporate purpose. For a long time, the adoption of new technologies was driven almost exclusively by efficiency metrics and cost reduction. What has become evident now is that the most admired companies — the ones with the greatest ability to attract talent and investment — are those that can articulate why they are using artificial intelligence, not just how. The narrative of automation for the sake of automation is losing steam. In its place, an approach is growing that connects technology to real-world impact — whether in sustainability, inclusion, public health, or customer experience. Organizations that can translate their AI initiatives into stories of tangible impact are building a type of competitive advantage that goes far beyond the quarterly balance sheet.
Another trend that is impossible to ignore is the democratization of access to artificial intelligence through lower-cost devices. Until recently, running AI models required heavy investment in cloud infrastructure and highly specialized teams. Today, dedicated chips that fit in a smartwatch can execute complex inferences locally. This opens doors for small and medium-sized businesses that were previously completely shut out of this game. A family farmer with a drone equipped with AI can map crop pests with the same precision that a large agribusiness corporation achieved five years ago with equipment that cost ten times as much. This democratization is not just a technological trend — it is an economic and social force that will reshape entire supply chains in the coming years.
Finally, it is worth highlighting the growing movement toward open standards and interoperability between AI ecosystems. The major manufacturers of smart devices and the leading artificial intelligence platforms have begun to realize that walled gardens limit value creation for both themselves and their customers. Standardization initiatives that allow AI agents from different vendors to communicate with each other, share context, and collaborate on complex tasks are gaining real traction. For enterprise security, this means more complexity in governance, but also the possibility of building smarter and more coordinated defenses. For companies overall, it means the freedom to assemble technology stacks that truly make sense for their needs, without being locked into a single vendor.
What to expect in the months ahead
The landscape taking shape for the coming months points to acceleration across all of these fronts simultaneously. NIST standardization for agentic AI is expected to produce its first formal documents before the end of 2025, which will create a global reference for autonomous agent governance. Smart devices with embedded AI processing should become more affordable and more powerful with each quarter, driven by fierce competition among chipmakers. And the conversation about corporate purpose in the age of AI will keep gaining ground in boardrooms and strategy sessions around the world.
The future of enterprise AI is open, distributed, and deeply integrated with the physical world — and it is arriving faster than most people realize. Organizations that see this moment as an opportunity to rethink not just their processes but also their reason for existing will be in the best position to thrive in this new cycle. 🔮
