My take: Mythos rejuvenated the cybersecurity sector. Earnings put the recent rally to the test

The Mythos bounce was always going to hit a wall eventually
I enjoy watching markets react to cybersecurity earnings because the gap between what actually happened and what investors wanted to happen is usually revealing. This week it was stark. CrowdStrike and Palo Alto Networks both posted strong numbers, both gave accelerating guidance, and both lost ground on the day. CrowdStrike dropped 8%, Palo Alto 3%. The results weren't the problem. The problem was the story investors had already written in their heads.
Anthropic's Mythos model changed the conversation for cybersecurity firms in an unequivocal way. A model powerful enough that Anthropic initially deemed it too dangerous to release publicly because of how easily it could be leveraged to exploit software vulnerabilities is the kind of thing that makes enterprises suddenly very interested in their security posture. CrowdStrike and Palo Alto were both early partners in Anthropic's Project Glasswing testing program. Shares of both companies ran more than 70% between April and the end of May. That is not a modest re-rating.
So when earnings arrived as the first real test of that rally, the bar was already sitting somewhere unreasonable. Palo Alto CEO150 said over 1,200 companies had reached out to discuss AI strategy, and that his team had held 800 meetings in six weeks, close to 100 of which he conducted personally. That is a real demand signal. But he was also clear that investors shouldn't expect a sudden windfall next quarter.
"I wouldn't get ahead of my skis and start throwing the kitchen sink and numbers for cybersecurity companies, because there is still a process, a mechanism, a cycle that people buy in, and there's execution and deployment."
That's a CEO being honest in a way that Wall Street doesn't always reward. Enterprise sales cycles in this sector typically run nine to twelve months. If a company only launched an AI-adjacent product in the last quarter or two, the revenue from that doesn't materialise immediately. Jefferies analyst Joseph Gallo pointed out that most of the AI uplift probably won't show up in numbers until calendar year 2027, with Q4 being the natural buying season as enterprise budgets reset. That's just how the procurement cycle works.
What's interesting is that the source article frames this as investors being impatient, which is fair, but it skips over a legitimate question: how much of the Mythos-driven rally was pricing in genuine future revenue, and how much was just sector sentiment catching a bid? A 70% move in six weeks for mature cybersecurity companies is not a conservative forward-looking estimate. It derives more from narrative momentum than from discounted cash flow. CrowdStrike's George Kurtz described AI detection and response as a potential market that could dwarf endpoint security, but also said it's in the "early innings." Those two things cannot both justify a 70% run at the same time.
The threat model shift is real, even if the timing isn't
The underlying logic for why Mythos matters to cybersecurity is sound and forward-looking. A model capable of identifying and exploiting vulnerabilities at scale is definitely an accelerant for attackers. Defenders need to move faster. That's a durable structural argument for the sector, not a one-quarter story. CrowdStrike lifting its fiscal 2027 net new ARR growth guidance on AI tailwinds is consistent with that.
But the market was demanding proof this quarter, which was always unrealistic given when Mythos launched and how enterprise buying actually works. Understandable that investors got excited. Inevitable that the first earnings test would be a disappointment against that setup. The more interesting data point will be six months from now, when those 800 meetings Arora described start converting into signed contracts and actual deployments. That's when the Mythos thesis either holds up or it doesn't. Right now we're still in the part of the cycle where nobody actually knows."
Our Briefing
Weekly signal. No noise. Built for founders, operators, and AI-curious professionals.
No spam. Unsubscribe any time.