Organizations around the world are accelerating the deployment of artificial intelligence (AI) applications and building AI-prepared data centers to promote automation and improve operational efficiency. This requires an AI reasoning environment and large language models (LLM) to be extremely efficient in processing real-time feedback, while also requiring new cybersecurity solutions to comprehensively protect the system.
Prevent, detect and mitigate cybersecurity threats A10 announces new AI fireworks features, which can be deployed to protect API's or URLs connecting to language models. Built on an edge-optimized architecture and integrated GPU hardware, these features help protect LLM with high performance and can be flexibly deployed as an additional layer of security in all existing infrastructure.
Accordingly, the system can detect threats such as malicious injections and sensitive information leaks by monitoring request-responsive flow at the prompt level and implementing necessary security policies to minimize risks. Bringing real-time experience A10 continues to enhance performance and sustainability for AI and LLM- integrated applications.
Accordingly, new capabilities also allow early detection of network performance problems similar to an early warning system to help identify network congestion risks or resource shortages in the near future...
Enterprises are rapidly deploying and training AI and LLM inference models both on-site and in the cloud. To meet the three main challenges of these new environments latency, security and operational complexity , new appropriate capabilities are needed. With more than 20 years of experience in security and application distribution, A10 is expanding solutions to meet those needs, to bring resilience, high performance and safety to AI and LLM infrastructure, said Dhrupad Trivedi, Chairman and CEO of A10 Networks.
