Flash News

CrowdStrike CEO: AI is being built into every company, every industry, every country; the future requires a parallel of cutting-edge closed-source and open-source models

CrowdStrike CEO George Kurtz stated that AI is being built into every company, every industry, and every country. In the future, there will be a need for cutting-edge closed-source and open-source models to run in parallel, with open-source models providing a security advantage.

He pointed out that more teams are inspecting, testing, and reinforcing models, allowing defenders to use capabilities equivalent to those of attackers, and proudly noted that CrowdStrike is the only cybersecurity company to sign this support letter.

The discussion around event-driven cybersecurity and open-source AI is heating up, with funding flowing towards research on the security of open-source models and defensive tools, benefiting cybersecurity vendors and the open-source ecosystem, while putting pressure on purely closed-source dependency paths.

Source: Public information

ABAB AI Insight

As the founder and CEO of CrowdStrike, George Kurtz has previously positioned the company as the "pick and shovel" of cybersecurity in the AI era, continuously emphasizing exposure management and platform-based defense. His public support for open-source weighted models continues his assertion of parity in capabilities for defenders.

On the capital front, CrowdStrike stands with open-source by signing the joint letter, motivated by the need to ensure defensive tools can access model capabilities equivalent to those of attackers, while solidifying its unique position in AI security infrastructure.

In comparison to Microsoft, NVIDIA, and others who signed the same letter, AI security is currently transitioning from a closed model dominance to a phase of parallel validation of open-source and closed-source.

This essentially represents a technological substitution: open-source models replace single closed-source trust through multi-party verification, with the mechanism being to expand the review scope to enhance overall security, reallocating defensive resources from passive response to proactive hardening.

ABAB News · Cognitive Laws

  1. Defense must have weapons equivalent to attack.
  2. The true value of open-source is to allow more eyes to inspect.
  3. Security is a prerequisite for the scaling of AI, not a patch after the fact.

Source

·ABAB News
·
3 min read
·1d ago
分享: