It is not new that cyber attacks have become more precise and sophisticated; It seems that forecasts will become more complex in the future. According to “Cyber Threat Forecasting Report 2026“L FortniteCybercrime will continue to evolve into a regulated industry, building automation, specialization, and artificial intelligence (AI).
“Thanks to AI, automation, and a unified cybercrime supply chain, intrusions will become easier and more resilient than ever before. Attackers will spend less time developing new tools and more time mastering and automating already effective technologies. AI systems will manage reconnaissance, accelerate intrusions, analyze stolen data, and manage ransom negotiations. At the same time, independent cybercrime agents on the dark web will begin executing full stages of attacks. Attacks with minimal human oversight,” this report explains.
In the face of this complex scenario, it is important for companies to be prepared. Specifically, with the aim of knowing the situation of local companies, Microsoft You conducted your own survey Cybersecurity 2025: Challenges and strategies in the era of artificial intelligence for large companiescommissioned by Edelman.
The survey, which includes the views of 100 security and IT professionals in Argentina, provides an updated view on how organizations are adapting their approaches to deal with changes in the digital environment.
Among the main findings stands out Six out of ten IT professionals consider cyber threats a specific challenge For their organizations, in a context of increasing digital risks. 76% believe that these threats have increased In the last two or three years and 78% expect cybersecurity to remain a priority in their companies during the same period.
Among the risks most mentioned by organizations are: Malware (malware) by 53% and Phishing (Attacks designed to deceive users and obtain confidential information) by 42%.
The good news is that in this scenario, artificial intelligence (AI) is starting to take center stage in organizations’ response. 95% of companies consider AI to be a factor of high or moderate impact In cybersecurity practices, by enhancing defense against increasingly sophisticated threats. While currently approx Five out of ten companies reported a high or moderate reliance on AI To address cybersecurity risks.
“Today, AI is redefining how we work, innovate and protect the things that matter most,” says Sebastian Avil, Country Director of Microsoft Argentina. “At Microsoft, we believe that trust is the foundation of digitalization, which is why security, privacy and responsible use are our priorities. Today, AI is not only increasing productivity, but also becoming the greatest ally in strengthening cybersecurity in the face of an increasingly complex threat scenario.”
Another encouraging fact revealed by the report of the company founded by Bill Gates is that 74% of Argentine companies have a specific cybersecurity strategy and 88% indicate that this component is integrated into the overall business strategy.
But there is still a long way to go, as a result of these developments alone 35% of the companies included in the study consider themselves highly prepared to confront current threats, While 47% had an average preparation level. Key strengths include data protection (53%) and cybersecurity policies (48%), while identity protection (20%) and employee training (26%) still need to be strengthened.
the The new main champions of cybersecurity appear to be artificial intelligence agents. They are increasingly chosen for this task due to their great ability to optimize critical processes and improve operational efficiency in various fields.
Five out of ten companies have formal policies for the use of AI agents, and eight out of ten consider these guidelines to be effective. “Although the perception of effectiveness varies depending on the type of organization, 33% of digital-native companies rate these policies as highly effective, while in non-digital-native companies this level is only 29%,” the report explains.
In this sense, the majority of respondents (58%) reveal that they are already using AI agents in cybersecurity operations and 37% are evaluating their integration, reflecting the gradual progress in their adoption.
But this integration is not limited to the field of security AI agents also have a presence in other areas of operations: IT infrastructure records 59% of usage and the financial and administrative field reaches 52%. These numbers show their integration into jobs that require speed of movement and continuous management, which in turn contributes to enhancing digital security operations.
According to a Microsoft survey, the contribution of these solutions appears to enhance key tasks. Agents, designed to automate activities, coordinate flows, and act under safeguards, facilitate reduced investigation times, prioritization of alerts, incident containment, and consistent enforcement of identity and data policies. The combination of AI models, large-scale telemetry, and Zero Trust-based controls helps strengthen the security posture against increasingly complex threats.
“AI not only increases the speed and scale of defenses,” says Marcelo Fellman, director of Microsoft Cybersecurity for Latin America. “With specialized agents, it helps close talent gaps, standardize processes, and increase business agility. The challenge is no longer ‘if’ to invest in security, but ‘how’ to scale it with AI and responsible governance at every layer of the organization.”
Microsoft offers six recommendations when implementing cybersecurity strategies in an increasingly complex environment:
1. Make of Safety is a priorityRisk management from management level.
2. Promote good digital habits. It’s important to prioritize identity protection, secure the enterprise perimeter, and continually monitor cloud assets.
3. Enhancing digital resilience. It is important to identify vulnerabilities, plan ahead, and train employees to respond and recover quickly from any breach.
4. Invest in peopleAnd not just in tools.
5. Prepare for the future. Anticipate changes, understand that “we don’t know what we don’t know,” and always share threat information.
6. Make the easiest decision and the right decision. Instead of banning and banning, encourage safe use.
“There’s already been talk about ‘shadow AI’ in reference to employees using AI without their employer giving them the tool,” Fellman acknowledges. “The problem is that the IT area has no visibility or control over what these people are doing. It could, for example, send the customer base to the AI to create an automated campaign. What happens if someone blocks access from that employee? The answer is I lost.”
For Microsoft’s Cybersecurity Director for Latin America, “It’s no different than what we’ve done in traditional security. The best antidote is to make it easier for employees to make the right decision. 99% of attacks can be avoided by following these recommendations.”