Mythos Boosts Cybersecurity, Earnings Test Rally

Anthropic’s Mythos model has emerged as a potential game-changer for cybersecurity firms navigating the rapidly evolving landscape of artificial intelligence.

However, this past week’s cybersecurity earnings reports served as a stark reminder that even with significant tailwinds, strong performance isn’t always enough to satisfy market expectations. Shares of CrowdStrike and Palo Alto Networks saw notable declines of 8% and 3%, respectively.

“Investors may have gotten a bit ahead of themselves,” commented Joseph Gallo, a software analyst at Jefferies. “While both companies provided optimistic guidance, the reality is that realizing the full benefits of AI integration is a multi-year process and takes time to manifest in financial results.”

Earlier in the year, cybersecurity stocks experienced a downturn amid concerns that new AI tools, capable of accelerating software development at an unprecedented pace, could disrupt their established business models and, by extension, the broader software industry.

The subsequent unveiling of Mythos, a powerful AI model initially deemed too potent for public release due to its potential for exploiting software vulnerabilities, reignited enthusiasm for the sector. This renewed optimism fueled a significant rally, with shares of CrowdStrike and Palo Alto Networks each climbing over 70% between April and the end of May.

Both companies were early participants in Anthropic’s exclusive Project Glasswing testing program. Anthropic recently expanded this initiative to include 150 additional partners, a roster that now features prominent players like Rubrik and Tenable.

This quarter’s earnings reports represented the first significant market test for the Mythos-driven rally. Despite delivering strong results and optimistic commentary regarding AI’s potential, neither cybersecurity giant was able to fully assuage investors demanding immediate tangible returns from AI investments.

The principle of “good not being good enough” is a familiar narrative on Wall Street. Even market darling Nvidia, a prominent beneficiary of the AI boom, has experienced pullbacks after failing to consistently exceed exceptionally high investor expectations.

Investors approached this earnings season with a similar cautious optimism, hoping that AI-driven tailwinds would propel cybersecurity firms to deliver outstanding financial results.

While the reported earnings contained elements for analysts to appreciate, investors may be overlooking the crucial point that these AI tailwinds could take several months to translate into significant financial gains, according to Gallo.

Typical enterprise sales cycles span nine to 12 months, suggesting that the substantial impact of AI adoption is unlikely to be fully reflected in financial reports until the 2027 calendar year. Furthermore, the fourth quarter of the calendar year is traditionally the strongest period for enterprise spending as businesses finalize budgets for the upcoming year.

“If an enterprise has only recently launched an AI product in the last quarter or two, it’s unrealistic to expect a massive immediate uptick in demand and revenue,” Gallo added.

The CEOs of the leading cybersecurity companies themselves have underscored this point.

Palo Alto Networks CEO Nikesh Arora informed analysts this week that demand in the Mythos era is “off the charts,” with over 1,200 companies reaching out to the firm to discuss their AI strategies. Arora indicated that Palo Alto Networks has conducted approximately 800 meetings over the past six weeks, with a significant portion of those involving him personally.

While demand patterns are showing positive indicators, Arora cautioned that analysts should not anticipate an immediate “windfall” in the next quarter as businesses integrate AI into their cybersecurity frameworks. However, he expressed confidence in “robust growth” moving forward.

“I wouldn’t get ahead of ourselves and start projecting overly aggressive numbers for cybersecurity companies just yet, as there is still a well-defined process, a procurement mechanism, and a deployment cycle that needs to be completed,” Arora stated.

CrowdStrike CEO George Kurtz echoed this sentiment, highlighting the long-term nature of AI integration.

The company has raised its fiscal 2027 net new annual recurring revenue growth targets, attributing the upward revision to anticipated AI tailwinds.

Kurtz explained to analysts during an earnings call that AI detection and response (AIDR) represents a substantial new market segment that could potentially surpass the current endpoint security market. He emphasized, however, that this segment is still in its “early innings.” CrowdStrike’s pipeline for the second quarter has already exceeded $50 million, Kurtz noted.

“Once AIDR truly becomes mainstream and entire organizations adopt it across all their employees and workloads, I believe you will see a significant increase in incremental opportunities,” Kurtz projected.