A10 Networks, a leading global provider of secure application and network infrastructure solutions, announced it has completed the acquisition of TrojAI Inc., an innovator in enterprise artificial intelligence security.
The strategic acquisition accelerates A10’s corporate technology roadmap, embedding multi-layer AI application governance, build-time risk testing, and sovereign runtime threat defenses into its global networking security portfolio. The transaction allows enterprise organizations and public-sector authorities to maintain complete control over where, when, and how their proprietary AI models, training datasets, and autonomous agents are shielded from adversarial exploits.
Neutralizing the Non-Deterministic Attack Surface
The enterprise integration of Generative AI and autonomous software agents has introduced complex security vulnerabilities that standard, deterministic firewalls are unequipped to handle. Large Language Models (LLMs) suffer from unique vulnerabilities-including prompt injection bypasses, training data poisoning, and training model extraction attacks-that exploit the core probabilistic nature of AI processing rather than traditional software code flaws.
Also Read: TrendAI Integrates Anthropic’s Claude Compliance API into Vision One Platform
These vulnerabilities can be combatted with TrojAI, which is designed with an all-inclusive dual layer security system built specifically to provide protection for the AI pipelines within corporations throughout their entirety:
Build-time Red Teaming: An automatic security system that stress-tests, red-teams, and searches for vulnerabilities within LLMs and customized agentic processes at build time, revealing weaknesses in the models prior to any deployment.
Runtime Security: A latency-focused security protocol that watches for any inference requests made to the systems, screening both input and output for any possible issues or threats.
“AI is changing both what enterprises build and the attack surface they have to defend, and traditional controls weren’t designed for non-deterministic models and autonomous agents,” said Dhrupad Trivedi, President and Chief Executive Officer of A10 Networks. “TrojAI is a natural fit for A10, strategically and operationally. Pairing our hardware-based AI firewall with TrojAI’s software-based red teaming and runtime protection helps customers adopt AI quickly and confidently, protecting their models, data, and agents without sacrificing the latency or availability they rely on us for, whether on-premises, in the cloud, or hybrid. For customers with strict data-sovereignty requirements, it means embracing AI while keeping their most sensitive assets in environments they control.”
Consolidating Hardware Firewalls with Cloud-Native Software Defenses
By integrating TrojAI’s software assets into its portfolio, A10 Networks bridges the gap between hardware-accelerated perimeter protection and deep, app-layer algorithmic security. This combined offering meets the demands of highly regulated sectors-such as financial services, defense, and healthcare-where strict international data residency laws require companies to isolate sensitive processing workloads inside local data centers or hybrid environments they completely manage.
The engineering integration will allow IT operations teams to deploy unified policies that monitor data flows from the network backbone down to separate, isolated LLM endpoints, keeping corporate assets secure regardless of underlying cloud storage setups.
“Enterprises and public-sector organizations are adopting AI at an unprecedented pace, and they need to innovate securely while maintaining sovereignty over their AI security infrastructure,” said Lee Weiner, Chief Executive Officer of TrojAI. “Together with A10, we can secure and govern the models, agents, and applications becoming core to how organizations operate. I’m proud of what our team has built, and excited to bring these capabilities to A10’s customers and channels.”