Cybersecurity Stocks in Panic Selloff—But the AI Threat Is Far From Imminent

The selloff in cybersecurity stocks is a direct reaction to a specific, leaked event. Last week, internal documents from AI firm Anthropic surfaced online, revealing the company is testing a model it calls Claude Mythos. According to the leaked draft, this is the most powerful AI model it has ever built, introducing a new tier above its existing Opus line. The core threat narrative is clear: Mythos is being developed with "meaningful advances in reasoning, coding, and cybersecurity," and Anthropic itself warns the model could significantly heighten cybersecurity risks by rapidly finding and exploiting software vulnerabilities.

This is not a finished product. The model is still in an early phase, currently being tested with a small group of early access customers as Anthropic evaluates its behavior and risks. The company has explicitly stated it is being "deliberate about how we release it," indicating a cautious, non-immediate rollout. Yet, the market interpreted the leak as a near-term existential threat to the entire sector.

The tactical overreaction was swift and sharp. On Friday, shares of key cybersecurity players fell on the news, with Palo Alto Networks (PANW) stock falling an unlucky 7%. CrowdStrike and Fortinet also dropped 4% to 6%, dragging down the broader tech-software sector. The sell-off was driven by fears that a model this advanced could render traditional cybersecurity tools obsolete, threatening market share and pricing power. In reality, the event creates a potential mispricing: the threat is real and long-term, but the immediate catalyst is a model that isn't even publicly available.

The Immediate Reaction: Panic vs. Fundamentals

The price action tells a clear story of narrative-driven panic. CrowdStrike shares are down 9.6% over the past five days and have fallen 24.6% over the past 120 days, trading near their 52-week low of $298. This isn't a new selloff; it's a continuation of a deep, multi-month decline that has already punished the stock. The recent drop on the Anthropic leak is a sharp spike on top of a pre-existing downtrend, suggesting sentiment is amplifying a broader fear rather than revealing a sudden, fundamental deterioration.

Analysts point directly to the disconnect. Shrenik Kothari of Baird called the sell-off "a panic-driven, narrative-led selloff". The core fear-that AI could render existing tools obsolete-is simplistic and ignores how security works in practice. As Kothari noted, Anthropic's current tool, Claude Code Security, does not handle real-time security tasks like stopping live intrusions. The market is pricing in a future threat while ignoring the present-day reality of how these platforms are deployed and generate cash.

This sets up a critical valuation risk. Palo Alto Networks, a bellwether, trades at 35 times trailing free cash flow. That multiple leaves almost no margin for error on any perceived threat. When a stock is priced for perfection, even a speculative narrative about a future product can trigger a sharp sell-off. The event creates a mispricing: the long-term AI threat is real, but the immediate catalyst is a model that isn't even publicly available. For now, the market is reacting to the headline, not the fundamentals.

The Threat Assessment: Real but Distant

The leaked AI capabilities do represent a genuine, long-term threat, but it is not an immediate existential one. The internal documents warn the new model could significantly heighten cybersecurity risks by rapidly finding and exploiting software vulnerabilities, potentially accelerating a cyber arms race. This is the core of the market's fear: a new class of AI could outpace human defenders and traditional automated tools.

Yet, the practical reality is far more constrained. The model, dubbed Claude Mythos, is still in an early testing phase with a small group of customers. Anthropic itself is being deliberate about its release, and the company has stated it plans to share test results with cyber firms to help them improve defenses. As Bernstein analyst Peter Weed noted, this is "good hygiene and a baseline expectation for their product." The model is not a public tool that hackers can immediately deploy.

This is where the CEO's defense becomes critical. CrowdStrike's George Kurtz has publicly stated the new AI tool will not replace its widely known Falcon system. His argument is that AI scanning code for vulnerabilities is a different function from running an independent, battle-tested platform to stop active breaches. This distinction is key. The market is pricing in a future where AI automates the entire security stack, but the current tools are more like advanced diagnostic aids.

The broader sector is already under severe pressure, which makes it vulnerable to this narrative. The Global X Cybersecurity ETF fell as much as 6.1% on the news, bringing its decline this year to more than 20%. This deep, pre-existing downtrend means the stock is trading on fear, not just fundamentals. The leak has amplified an existing selloff, creating a mispricing where the distant threat is being valued as if it were imminent.

The bottom line is one of timing. The threat is real and will evolve, but the catalyst is a model that isn't even publicly available. For now, the tactical setup is clear: the market is reacting to a headline, not a product.

The Setup: Risk/Reward and What to Watch

The tactical setup is defined by a clear mispricing and a short list of near-term catalysts that will either confirm the threat narrative or reveal it as overblown. The key risk is a sustained narrative-driven derating, but the technical setup shows extreme pessimism priced in.

First, watch for any official product announcements from Anthropic. The company has a scheduled enterprise briefing, and traders are watching for any pre-announcement or commentary from management ahead of the next earnings cycle. Traders are watching for any pre-announcement or commentary from management ahead of the company's next earnings cycle. This event could clarify the security tool's capabilities and release timeline, providing a concrete data point to assess the threat. Until then, the leak remains a speculative catalyst.

Second, monitor management commentary from cybersecurity firms ahead of next earnings. The recent selloff has amplified existing fears, but the real test will be what leaders say about AI integration plans. CrowdStrike's CEO has already stated the new AI tool will not replace its widely known Falcon system. Other firms will likely follow, and their framing of AI as a complementary diagnostic tool versus a disruptive replacement will be critical. Any hint of a shift in strategy or investment priorities could move the stock.

The risk/reward hinges on this narrative. The key risk is a sustained derating if the AI threat narrative gains traction, but the technical setup suggests the worst may already be priced in. CrowdStrike shares are down 9.6% over the past five days and have fallen 24.6% over the past 120 days, trading near their 52-week low. This deep pessimism creates a potential floor; further downside would require a fundamental deterioration, not just a narrative. For now, the event-driven selloff has created a mispricing where the distant threat is being valued as if it were imminent. The catalysts ahead will determine if this is a buying opportunity or the start of a longer correction.