A10 Networks, Inc

A recent report on application and API security makes one thing clear about the application security space: It's about implementing an integrated, intelligent, and seamless approach to protecting applications against attacks and attackers. Read more about the report's takeaways in Gary Wang's blog post at the link. https://bit.ly/3PqdrMw

Watch A10's Field CISO Jamison Utter and Redmondmag's Editor-in-Chief John K. Waters discuss how AI is reshaping how applications develop, deliver, and scale. The conversation covers evolving delivery pipelines, optimizing for performance and reliability, and practical strategies for accelerating innovation across cloud-native and legacy environments without sacrificing control. https://bit.ly/4ukDFyD

Business logic fraud is among the most damaging and difficult attack types for organizations to detect, because attackers operate entirely within the bounds of your own systems. A10's Gary Wang uses a real-world referral program fraud scenario to illustrate the difference between dedicated bot management platforms and ThreatX, and why the two approaches work best together as a layered defense. Learn more about ThreatX: https://bit.ly/3QVFiVb

As applications grow more complex and AI-driven, the infrastructure supporting them needs to keep pace. A recent IDC Technology Spotlight Report examines why AI-powered applications are accelerating demand for smarter API load balancing and adaptive infrastructure, how modern ADCs enable agility and scalability across hybrid and multicloud environments, and the key benefits of centralizing management and automating policy enforcement for distributed AI workloads. Get the report: https://bit.ly/4cRWPoD

As large language models become embedded in production systems across enterprises, the security architecture required to protect them looks fundamentally different from anything organizations have deployed before. In LLM environments, natural language becomes an execution vector. Attackers exploit interpretation rather than code vulnerabilities. A carefully crafted prompt can override system directives, surface sensitive data from connected knowledge bases, or influence downstream API calls and automated workflows. But traditional security tools built for deterministic code paths were not designed to catch any of that. This guide to LLM security in 2026 covers how to evaluate LLM security tools built for the inference layer, not just the perimeter. ➡️ https://bit.ly/4sO7c2y

Financial institutions are accelerating AI adoption, but doing so at a risk. The overall attack surface expands across APIs, web applications, and automated traffic. Now more than ever, financial IT teams need to know how to implement AI guardrails, modern app security controls, and intelligent load balancing to protect AI-enabled systems without slowing transactions. During this live webinar, John K. Waters of Redmondmag and A10's Jamison Utter will explore strategies to reduce risk, protect customer trust, and maintain visibility across increasingly complex digital ecosystems. Register today for May 27 at 1 pm ET: https://bit.ly/3PpzPW6

A10 had a prominent presence at CYBERSEC in Taipei, the largest cybersecurity trade show in the region. This year, we introduced in-booth tech talks. Over the three-day event, we successfully conducted 12 sessions covering AI firewall, DDoS protection, Cloud Access Proxy, and API security, featuring A10 speakers Nick Chen and Allen Lin alongside partner Unicomp.

As large language models (LLMs) become core to enterprise infrastructure, the risks they introduce are new. An AI firewall sits between applications and large language models, inspecting natural-language traffic and protecting infrastructure against prompt injection, jailbreak attacks, data poisoning, system prompt leakage, and OWASP Top 10 vulnerabilities. By keeping data sent to and received from LLMs safe, compliant, and high-quality, it serves as a critical buffer for both policy enforcement and organizational trust.

Generative AI adoption introduces critical security challenges, including prompt injection, data leakage, and malicious code generation. Organizations must implement layered defenses, such as AI firewalls and output filtering, to protect sensitive data and maintain model integrity. Integrating these specialized controls with infrastructure resilience ensures that AI workloads remain both secure and available during enterprise deployment.

Most organizations are still working through basic AI security challenges while pressure builds to prepare for more advanced threats. That gap often leads to AI security programs that look mature on paper but lack visibility and control in practice. Most enterprises are not building autonomous AI systems or training foundation models. They are connecting managed APIs to internal workflows, integrating LLMs into existing tools, and allowing employees to use platforms like ChatGPT because blocking adoption entirely was never realistic. The immediate risks are operational. Employees may upload sensitive data into tools they do not fully understand. Teams may create API keys with broad permissions and no expiration policies. Security teams are often involved only after AI tools are already embedded across the business. A strong AI security posture starts with knowing what is deployed, what data it can access, and how failures are detected and contained. Read more: https://bit.ly/48ApnSi