CrowdStrike and Meta have launched CyberSOCEval, an open-source benchmark suite designed to evaluate the performance of AI systems in security operations, focusing on real-world adversarial scenarios.

CyberSOCEval assesses large language models (LLMs) across critical security workflows such as incident response, malware analysis, and threat comprehension, building on Meta's CyberSecEval framework and incorporating CrowdStrike's threat intelligence.

This initiative aims to help security teams measure AI effectiveness against evolving threats, enabling organizations to identify the most valuable AI applications in their cybersecurity efforts.

Major analysts like Truist Securities and RBC Capital have reiterated buy ratings for CrowdStrike, with target prices around $500-$510, reflecting confidence in its growth despite some cautious near-term revenue outlooks.

Analysts remain optimistic about CrowdStrike’s future, with firms like Truist Securities, TD Cowen, UBS, and RBC issuing strong buy or outperform ratings, supported by its robust financial position.

The rise of AI has sparked a digital arms race in cybersecurity, with malicious actors using AI for attacks like password brute-forcing, while security teams deploy AI-powered defenses, saving millions in sectors like finance.

CrowdStrike's chief business officer, Daniel Bernard, highlighted that this partnership with Meta sets a new standard for AI deployment in cybersecurity, helping defenders stay ahead of increasingly sophisticated threats.

The initiative seeks to establish new standards for AI development and deployment in cybersecurity, enabling defenders to better understand and leverage AI capabilities.

This collaboration underscores a growing focus on integrating AI into cybersecurity to improve threat detection and response capabilities.

Many security teams are still early in their AI adoption, and without benchmarks, it’s difficult to determine which AI systems provide genuine advantages against cyber threats.

Cyber defenders face increasing challenges from rising security alerts and evolving threats, emphasizing the need for effective AI tools in Security Operations Centers.

The CyberSOCEval tool is designed to help security teams understand AI models' capabilities and limitations in real-world security scenarios.

These benchmarks assist security teams in identifying where AI adds the most value, providing targets for capability improvements and formalizing evaluation processes for AI in cybersecurity.

CrowdStrike has exceeded fiscal guidance across all metrics, with a forecast of over 40% growth in net new annual recurring revenue in the second half of fiscal year 2026, prompting an upward revision of its outlook.