GenAI adoption across security domains

As use and adoption of? Generative AI (GenAI) and Large Language Models (LLMs) continues to grow, securing GenAI and LLMs are of particular concern. Approaches to mitigate GenAI and LLM security, commonly referred to as “Securing AI” is an innovation hotbed and active area of VC investment.??

Use of GenAI and LLMs for security use cases, commonly referred to as “GenAI for Security” has reinvigorated the roadmap of 4,000+ security vendors, has served as a catalyst leading to? new startups as well as use of GenAI by security teams for internal use cases.

This blog shares a summary of the current state of “GenAI for Security” solutions across key cybersecurity domains. The solution status is a subjective view :-).?

Security Domains

Security domains cover areas critical for a robust cybersecurity strategy, including access control, network, governance and risk management, software supply chain, encryption and cryptography, data, privacy architecture and design, operations, physical and IOT, legal, training, regulations and compliance, and offensive security. Each domain addresses aspects of protecting information (or physical) systems, from managing user access and securing networks to ensuring compliance with legal standards and responding to security incidents. Together, these domains provide a framework for IT leaders to safeguard data, maintain system integrity, and protect against a wide range of threats and vulnerabilities.

Note: There are different security domain frameworks including one from NIST. Pick your favorite framework to assess its current state.

GenAI use across Security Domains

There is an ongoing tendency to play up use and adoption of GenAI. Security vendors are not immune to this. It is not always clear about the specifics of how GenAI is being used.

GenAI use and adoption across Security domains varies. Augmentation is the form of a Co-pilot, Assistant, Expert, etc., is the most common approach. In some domains such as Network Security, Identity and Security Operations GenAI use is native.?

Vendor Strategy: GenAI for Security?

Majority of Security vendors are planning to or have adopted a hybrid strategy with GenAI starting with augmentation and expanding their solutions with native capabilities.

Three established security vendors, 微软 , Palo Alto Networks and CrowdStrike have extensively discussed their approach to adopt GenAI. All three vendors have a hybrid strategy with?a copilot and native capabilities.

Microsoft has a hybrid strategy with a Copilot for Security augmentation and native use of GenAI across their security portfolio including Intune, Purview, Entra, Sentinel, XDR and Threat Intel.

Palo Alto Networks recently introduced its “Precision AI” approach to simplify security by secure design, improved defense and response.

CrowdStrike Falcon platform has a single agent approach with an augmentation “Charlotte AI” assistant and native embedding across its tools.

Summary

Established and emerging security solutions have implemented GenAI and are already reaping its benefits. This applies to almost all security domains.???

I have used a few tools and they do better as an ally or assistant now. The automation and scale benefits are realizable for well understood use cases. More about these and related topics in a subsequent blog.

Thoughts? Let us discuss.