Author：Wall Street CN

Anthropic is handing over a more powerful, undisclosed AI model to tech giants for testing in response to potential cybersecurity threats from advanced AI systems. This move has triggered an immediate market reaction, with cybersecurity stocks rising in response.

Tuesday, April 7thAnthropic announced the formation of an industry consortium called "Project Glasswing," which will grant companies such as Amazon, Apple, Microsoft, and Cisco access to its unpublished new model, Mythos, to investigate vulnerabilities in their own products and share findings with industry peers.

Concerned that hackers might use the model to launch cyberattacks, Anthropic stated that there are currently no plans to release Mythos to the public, and that security safeguards for the technology will be developed based on feedback from Project Glasswing.

Global cybersecurity stocks generally rose on Tuesday, buoyed by the news. The Global X Cybersecurity ETF rose 0.9%, marking its sixth consecutive day of gains.

Palo Alto Networks rose 4.9%, CrowdStrike climbed 6.2%, Zscaler gained 1.8%, and Fortinet advanced 1.7%. Analysts believe that Anthropic's move clearly validates the boosting effect of AI technology on demand for cybersecurity software.

The core logic of Project Glasswing is to identify and mitigate potential cybersecurity vulnerabilities before powerful AI models are more widely adopted. Participating companies will leverage the Mythos model to proactively uncover flaws in their own products and share the research findings within the industry.

Newton Cheng, who is in charge of cybersecurity for the Anthropic frontier red team, said:

We believe this is not just an Anthropic issue; it's an industry-wide problem that both private companies and governments need to address. What we do with Glasswing is give defenders a first-mover advantage.

Analysts believe this actually reflects the growing concerns within the technology industry.As AI models continue to improve their capabilities, criminals and hackers may use such tools to scan for vulnerabilities in source code and breach network defenses.

Anthropic's competitor, OpenAI, has also previously emphasized the growth of its model network capabilities and launched pilot projects aimed at prioritizing the "handing over of defenders" of the tools.

Anthropic stated that it has communicated with U.S. government officials regarding Mythos's security capabilities, but declined to disclose which specific agencies were involved..

Newton Cheng pointed out that the company's current work already involves collaboration with the U.S. Cybersecurity and Infrastructure Security Agency (CISA) and the National Institute of Standards and Technology (NIST).

Wall Street analysts generally interpret Anthropic's move as a positive signal for the cybersecurity industry, believing that this event redefines the relationship between AI companies and security vendors.

Jefferies Financial Group believes that this news is a "clear signal of cooperation rather than competition" for Palo Alto Networks and CrowdStrike, and indicates that "the threat environment is reaching a turning point," believing that these two platform companies "have good potential for excess returns in the emerging AI era."

Financial services firm Stephens stated that the news "verifies the severity of more powerful AI models in cybersecurity and the necessity for cybersecurity vendors to collaborate with key enterprise software providers to jointly defend against complex threats."

Stephens predicts that enterprises will increasingly rely on trusted cybersecurity providers to protect increasingly complex, heterogeneous AI environments.

Bernstein, a private wealth management firm, pointed out from another perspective that "Anthropic clearly places great importance on code security and the problem of its models being maliciously exploited," and believes that this "does not pose much of an obstacle for cybersecurity vendors, and may even be a tailwind."