AI Is Starting to Flex Its Network Security Muscles

Three important ways that AI is shaping the future of network security include:

1. Improving security decision-making
2. Making network protection autonomous
3. Helping security teams become more efficient

With cyberattacks increasing in frequency, scale, and sophistication, AI has an important role to play in helping identify vulnerabilities and optimizing security policies. Currently, threat modeling is a complex and time-consuming process that often involves some manual effort, limiting the ability of security teams to stay a step ahead of malicious actors.

Using advanced AI engines to ingest and rapidly analyze enormous volumes of network activity data enables far more rapid modeling of potential threats, creating a clear picture of the attack surface and recommending specific actions or policies to improve an organization’s security posture.

While this capability is already well established, new AI models offer exciting possibilities to further enhance security decision-making, including:

• Linguistic-Adaptive Retrieval-Augmentation (LARA) — LARA is an emerging model that can help large language models (LLMs) better understand the context of an interaction to more effectively classify intent. With obvious applications for chatbots and other multiturn conversations, it may also give security teams a powerful tool for assessing malicious intent. 

• Retrieval-Augmented Generation (RAG) — RAG is a promising model designed to improve LLM output by referencing an authoritative or domain-specific information source in addition to its training dataset. This can result in greater accuracy and relevance of results and greater efficiency.

Both models offer great promise for helping security teams zero in on the most important vulnerabilities faster.

Although there are multiple things to do to improve network security, it is critical to know which ones should be done immediately. AI can ostensibly perform that role faster and more effectively than a team of human experts, empowering those experts to make the right moves right now.

The emergence of agentic AI — artificial intelligence that can act autonomously to achieve specific objectives — will transform critical security functions. Trained on LLMs that are optimized for network security, agentic AI tools have the potential to gain sufficient trust to perform certain functions without human intervention.

For example, an AI system can be trained to analyze network activity in real time, producing a risk score for certain behaviors. When the risk score reaches a predetermined level, the AI system will act autonomously to mitigate the risk, blocking the suspect traffic and notifying the security team of the action.

With attacks coming fast and furious, security teams can be easily overwhelmed. In the time it takes human experts to analyze the activity, the attacker could steal sensitive data and be long gone.

Although agentic AI is still being refined, it holds tremendous promise for network security. The ability to rapidly recognize and respond autonomously to a threat will likely become increasingly common in the years ahead.

AI is already improving the efficiency of network security operations in a variety of important ways. For example, AI-enabled chatbots can be used to normalize language between application owners and security teams. This bridges a key communication gap between two groups that speak very different languages, improving collaboration to strengthen security.

AI can assess devices across the network to determine the specific function of each device and recommend the appropriate security policies and network segmentation schemes. This dramatically reduces security team workloads while helping ensure that optimal security configurations are tailored to the unique contours of the environment.

As AI tools continue to mature, expect to see even more sophisticated uses to enhance operational efficiency, freeing more time for security professionals to focus on higher-level strategic tasks.

Akamai has taken a thoughtful approach by incorporating AI technologies into our solutions in ways that make a real difference for security professionals. 

For example, in the area of microsegmentation, we are using AI to automate the labeling of network elements, identifying the role of each, recommending appropriate security policies, and enabling “one-click policy.” We also deploy AI to do deep data mining to seek out potential vulnerabilities that otherwise go unnoticed.

And that’s just the beginning. Akamai will continue to explore emerging AI and related technologies to deliver the capabilities our customers need to strengthen their security posture.

AI is reshaping many aspects of our lives, from online shopping to content creation and artistic expression. So it’s not surprising that AI will play an expanding role in network security.

Advanced AI tools that are engineered to help protect against rapidly evolving threats will be essential building blocks for network protection strategies in the years ahead. The key is to select the tools that complement human expertise in ways that dramatically improve your security posture.

Your adversaries are already taking full advantage of AI to detect and exploit your infrastructure’s vulnerabilities. It’s time to start building your AI defenses for the future.