A10 Networks buys TrojAI to boost AI security tools

Acquisition overview

A10 Networks has acquired AI security company TrojAI, extending its push into security tools for AI models, applications and agents.

TrojAI's technology will be folded into A10's wider security portfolio, adding AI testing and live threat protection software to its existing products. A10 said the acquisition would not have a material impact on its financial results for fiscal 2026.

Agentic AI focus

TrojAI focuses on securing and governing AI applications and so-called agentic workflows, where software agents act with greater autonomy. Its products are designed to test AI systems before deployment and monitor threats once they are running.

Dual-layer security

A10 said the combination gives it two layers of AI security. One is red teaming, which probes models, agents and applications for weaknesses during development. The other is runtime protection, intended to defend systems after they go into production.

The transaction adds to a crowded AI security market as companies and public bodies move generative AI tools from experiments into operational systems. That shift has increased scrutiny of model behaviour, data handling and the risks tied to autonomous software agents.

A10 executives framed the purchase as an extension of an existing strategy rather than a change in direction. The company said it has been building that approach for the past two years as data centre architectures evolve and AI systems are deployed at larger scale.

Among the assets A10 highlighted was TrojAI's support for the Model Context Protocol, a developing standard linked to agentic AI workflows. A10 said that support would help it meet customer demand in a fast-emerging area.

A10 also pointed to a feedback loop between TrojAI's testing tools and its own guardrail models. According to the company, findings from red teaming can be used to update model defences in near real time based on threats observed in production environments.

A10 has an established business in application delivery controllers, distributed denial-of-service protection, application security and API security. It said adding TrojAI would broaden that offering by tying AI-specific controls to the rest of its network and application security products.

Enterprise reach

According to A10, TrojAI brings a customer base that includes Fortune 50 organisations. The company said that track record reflected demand for tools that can test and govern AI systems in large enterprises and public-sector settings.

Dhrupad Trivedi, President and Chief Executive Officer of A10 Networks, described the deal as a response to changing security requirements around AI.

"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.

He said A10 sees TrojAI's products as complementary to its own AI firewall and broader infrastructure footprint.

"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," said Trivedi.

That emphasis on sovereignty runs through A10's message on the deal. The company said customers want tighter control over where AI models, data and agents are protected, particularly when sensitive information must remain within specific environments.

For TrojAI, the acquisition gives its technology access to A10's larger customer base, global operations and channel relationships. TrojAI's management said that would help extend the reach of its products beyond what it could achieve alone.

Lee Weiner, Chief Executive Officer of TrojAI, said organisations are under pressure to secure AI systems while keeping control of their infrastructure.

"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.

He added that the combination would place TrojAI's tools within a broader platform used by large organisations.

"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," said Weiner.

A10 said the combined business would target security across AI, applications, APIs, DDoS defence and application delivery, aiming to give customers one platform spanning both traditional digital services and newer AI-driven workloads.