Jufuri: The market's fear of AI disrupting Internet security is excessive. Identity authentication and terminal security are actually "immune areas", selling creates buying opportunities.

Jefferies Financial Group Inc. (Jefferies) released a research report stating that despite the market's fear of potential disruption caused by large language models (LLMs) such as Anthropic, the core areas of identity authentication, network security, and endpoint security have shown strong resilience in the AI wave, essentially unaffected by the potential disruptive effects of artificial intelligence. Analyst Joseph Gallo emphasized in the report that the recent collective sell-off by investors in the cybersecurity sector reflects an excessive concern in the market about AI substitution effects, while ignoring the fundamental difference between "deterministic requirements" and "probabilistic generation" in security defense. The analyst team led by Joseph Gallo stated, "We acknowledge that the advancement of LLM technology does pose risks to cybersecurity. However, we believe that this risk is mainly concentrated in cybersecurity subsectors already recognized as risky (such as threat intelligence, code scanning, vulnerability management, security operations automation and security operation centers), rather than affecting identity verification, network protection, endpoint security, and other areas that we believe are more resistant to risk - this differential risk distribution makes the recent widespread sell-off in the market seem particularly unexpected. As for the most influential controversy, it still revolves around whether LLM technology will divert incremental cybersecurity budgets that enterprises should have invested; however, according to our previous research data, the risks faced by cybersecurity budgets during cross-domain redistribution have actually been significantly weakened." It is understood that the trigger for this market turmoil mainly stems from the exceptional reasoning and programming capabilities demonstrated by Anthropic's latest model, Claude Mythos, which has sparked investors' apprehension about the possibility of traditional security software being directly replaced by highly intelligent AI tools. Influenced by such sentiments, industry giants such as CrowdStrike (CRWD.US), Palo Alto Networks (PANW.US), Fortinet (FTNT.US), and SentinelOne (S.US) have all recently experienced significant pullbacks in their stock prices. However, Jefferies Financial Group Inc.'s in-depth research shows that this "substitution theory" is difficult to sustain in the short term, as LLMs are fundamentally tools based on probabilistic predictions, while key infrastructures such as identity verification and endpoint defense require extremely precise "black or white" judgments. In environments with very low fault tolerance, such as real-time blocking of unauthorized intrusions or verification of user permissions, existing deterministic architectures are still relied upon, rather than the fuzzy inferences that AI may bring. In terms of the volatility of specific market segments, Jefferies Financial Group Inc. categorizes the field of network security into "immune zones" and "risk zones." Areas such as threat intelligence analysis, static code scanning, and vulnerability management, which heavily rely on pattern recognition and processing by Beijing Vastdata Technology, are indeed more susceptible to direct impact from LLMs, as these are the strengths of generative AI. In contrast, identity security vendors represented by Okta (OKTA.US), as well as the network and endpoint defense frontiers led by Zscaler (ZS.US) and Palo Alto Networks, form natural technological barriers due to their involvement in complex underlying protocol interactions and hardware integration. Jefferies Financial Group Inc. believes that the role played by LLMs in these areas is more as an "enhancer" than a "disruptor", assisting security operations personnel in shortening response times through AI assistants (Copilots), rather than replacing existing defense mechanisms. Additionally, Jefferies Financial Group Inc. emphasizes that AI laboratories like Anthropic are providing some of their models to security companies in advance, indicating a cooperative rather than competitive relationship. Gallo stated, "Anthropic's documents indicate that this model will be able to exploit vulnerabilities in a way that far exceeds the defenders' capabilities, and Anthropic appears to be releasing models to cybersecurity vendors in advance, giving them time to strengthen their code bases and enhance their defense capabilities. We believe this is a sign of cooperation rather than competition, and it has positive implications for overall cybersecurity needs." Gallo added, "Although we understand that the increased uncertainty surrounding these headlines has created a market environment of 'shoot first, ask questions later', we believe that the improvements in models by Anthropic and other LLM companies are not surprising (this trend will continue). As the market has taken a somewhat blanket diagnostic approach to this risk, we believe that companies with strong fundamentals, a competitive advantage in the field of artificial intelligence (we are bullish on identity verification and network security/endpoint security), and reasonable valuations offer increasingly attractive investment opportunities." As of Monday's close of the U.S. stock market, several cybersecurity companies' stocks have shown an upward trend. Specifically, CrowdStrike closed with a 2.84% increase, Palo Alto Networks rose by 4.99%, and Fortinet also increased by 1.09%.