As the adoption of LLMs and Agentic AI accelerates digital transformation, companies are facing unprecedented automated threats. In this article, Arie Simchis, Regional Director for Central America and Latin America at Radware, analyzes why traditional cybersecurity approaches have become obsolete and explains the urgency of adopting unified strategies to ensure operational resilience without sacrificing the customer experience.
In today’s hyperconnected economy, companies’ technological infrastructure has undergone a radical decentralization. As applications, application programming interfaces (APIs), and artificial intelligence-powered services become the absolute core of digital businesses, organizations face increasingly sophisticated threats. In this context, Arie Simchis, Regional Director for Central America and Latin America at Radware, warns of an unavoidable paradigm shift. “Applications have become the primary interface between organizations and their customers, which also makes them the primary attack surface,” the executive states.
Faced with distributed architectures, Radware’s response focuses on providing unified protection that covers hybrid and multicloud environments. According to Simchis, the challenge with APIs lies in the fact that they are highly dynamic components and are often difficult for security teams to fully inventory and monitor. To mitigate this risk, the company moves away from solutions based solely on traditional signatures and is committed to a proactive approach.
Radware uses language models (LLMs), machine learning, and behavioral analysis to protect the entire API lifecycle, detecting business logic abuse and anomalous activities in real time. In addition, the platform provides deep visibility to control the so-called “shadow APIs” (those that operate without IT oversight) and incorporates testing tools to assess API resilience before they move into production environments.
The challenge of Agentic AI and automated threats
The rapid integration of Artificial Intelligence into corporate processes is rewriting the rules of the game. Companies are accelerating their automation projects, but this innovation brings with it unprecedented vulnerabilities. Simchis is categorical regarding this phenomenon: “The adoption of LLM models and Agentic AI systems significantly expands the attack surface.” Among the main dangers threatening these technologies are prompt injection attacks, sensitive data leakage, AI model manipulation, and unauthorized access to critical business workflows.
The concept of “Agentic AI” adds an additional layer of defensive complexity. These environments present unique challenges because autonomous systems have the ability to make decisions, execute actions, and interact with multiple applications at machine speed. Since Agentic AI relies primarily on APIs for communication between its components and access to data, the protection strategy must evolve. Radware addresses this challenge by protecting the fundamental connections between AI models, users, and backend systems.
In addition to this is the proliferation of automated attacks and next-generation malicious bots. To combat them, Radware combines global threat intelligence with advanced behavioral analysis. Simchis highlights that the company’s research initiatives, such as ShadowLeak and ZombieAgent, are vital for identifying modern abuse patterns and adversary behaviors before they become widespread threats. Thanks to this continuous knowledge, organizations are able to detect account takeover, scraping, and credential stuffing techniques based on the intent behind the activity rather than on known signatures.
The obsolescence of the traditional WAF in the face of operational continuity
In highly regulated sectors such as banking, retail, telecommunications, and government, ensuring the uninterrupted availability of digital services is not only an IT requirement, but also a corporate mandate. In these industries, downtime caused by an attack translates into a business continuity issue and a severe risk to the brand’s reputation. In response to the increase in denial-of-service campaigns, Radware provides permanent, real-time DDoS protection capable of mitigating application-layer and volumetric attacks without interrupting the flow of legitimate user traffic.
One of the biggest strategic mistakes corporations make today is entrusting their security to previous-generation tools. On this point, the executive reflects: “The threat landscape has evolved far beyond what traditional WAFs (Web Application Firewalls) were designed to address.” Today’s cyberattacks target APIs and use artificial intelligence-driven techniques that require highly adaptive mechanisms. Therefore, organizations are compelled to adopt a comprehensive approach that unifies WAF protection, API security, bot management, and DDoS attack mitigation within a consolidated platform, abandoning reliance on isolated tools that provide only limited visibility.
The landscape in Latin America and the vision for the future
The Latin American market is at a turning point. Organizations across the region are accelerating their digital transformation processes and their migration to the cloud. However, they are facing cybercriminals who have also sophisticated their methods through automation. According to Radware’s experience in the region, although there is growing awareness of the importance of managing bots and securing APIs, a large portion of Latin American companies are still in the early stages of maturity regarding comprehensive protection strategies. The executive notes a strong regional interest in acquiring solutions that not only reduce operational complexity, but also increase infrastructure resilience.
A fundamental pillar of this modernization is not neglecting the customer experience. Cybersecurity should not generate friction or delays. Through intelligent traffic analysis and adaptive mitigation, Radware’s technology stops malicious activity while minimizing latency and false positives, factors that directly impact revenue and trust in the brand across digital businesses.
Looking ahead to the coming years, Simchis envisions a clear scenario: “The future of cybersecurity will increasingly focus on protecting applications, APIs, and artificial intelligence-powered ecosystems, as they constitute the core of modern digital operations.” For innovation and security to evolve together, his final message to corporations that are just beginning this journey is emphatic: “The most important step is to abandon fragmented approaches and move toward a unified strategy based on visibility, resilience, and continuous protection of the entire application ecosystem.”
How to take advantage of the challenges of the new digital era?
For partners looking to support their customers in protecting applications, APIs, and artificial intelligence-powered environments, Radware offers a portfolio designed to address the challenges of the new digital era through a unified cybersecurity strategy. Through Licencias OnLine, partners can access solutions, training, and specialized support to develop new business opportunities in this segment.
Learn more about Radware and its offering for the Latin American market here:
