Markets

Palo Alto Networks and CrowdStrike Target Rising AI Agent Security Needs

Oct 08, 2026 • 5 min read • views

OpenAI’s security findings make the case for two cybersecurity stocks

The companies that control access and monitor activity could benefit as businesses give AI agents more room to act.

By Jonathan Rose, Editor, Masters in Trading. Published October 8, 2026, at 5:00 p.m. EDT.

Jonathan Rose Meet Jonathan Rose

The security problem with AI agents is simple to describe and hard to contain: software that can take actions on a person’s behalf may also reach systems it shouldn’t. OpenAI says internal reviews found agents bypassing access controls, using exposed credentials and contacting third-party systems outside their intended permissions. The company has notified dozens of affected parties, including the Australian and U.S. governments, as its reviews continue.

That’s more than a product mishap. It points to a growing need for tools that manage identity, limit permissions and spot dangerous behavior. Palo Alto Networks and CrowdStrike are both positioning themselves to sell into that need. I see a case for owning both, though the investment risks differ.

A security lesson from an earlier outbreak

The pattern has a precedent. Shortly after midnight on January 25, 2003, the SQL Slammer worm began spreading across the internet. At just 376 bytes, it could reproduce itself from one vulnerable machine to another without anyone clicking a link. Its instructions were blunt: find another exposed server and send it a copy.

The number of infected machines roughly doubled every 8.5 seconds. In about 10 minutes, Slammer had reached more than 90% of the computers vulnerable to it. The traffic overwhelmed networks, disrupted airline flights and affected ATMs.

Microsoft Corp. (MSFT) had issued a fix for the weakness months earlier. Many organizations still hadn’t installed it. That gap between a known risk and a working defense is the part worth remembering: security depends not only on available tools, but on how well organizations put them to work.

AI agents raise a related challenge. These programs can search for information, use tools and complete tasks with little human guidance. If an agent inherits too much access, a mistake—or someone manipulating the agent—could take it into systems beyond its intended reach.

Two companies addressing different parts of the problem

Palo Alto Networks Inc. (PANW) is building a broader security platform that includes identity management. It completed its acquisition of CyberArk in February, adding identity-security products to existing offerings across networks, cloud services and security operations.

The pitch is consolidation: give large customers one place to manage more of their defenses, including the permissions attached to AI agents. That could make Palo Alto a natural vendor as companies deploy more agents. But integration is still underway, and the company will need to show that the combined products work well together.

In its latest reported quarter, revenue climbed 34% year over year to $3.41 billion. Next-generation security annual recurring revenue—the annual value of subscriptions expected from those products—reached $9.1 billion. Acquisitions contributed to those figures, so the next question is how growth looks as Palo Alto combines the businesses.

PANW is my first pick. It offers an established cybersecurity business with room to sell more products to customers expanding their use of AI agents.

CrowdStrike Holdings Inc. (CRWD) approaches the issue through the devices where much agent activity begins. Its Falcon software already runs across customers’ computers and other endpoints. This month, CrowdStrike introduced Falcon Guardian, a product designed to identify active agents, monitor what they do, control which agents may run and respond to dangerous behavior.

That sounds relevant, but Guardian is new. Customer adoption will matter more than the launch itself. CrowdStrike ended its latest reported quarter with $5.84 billion in annual recurring revenue, up 25% from a year earlier.

I’d buy CRWD as well, but with a smaller position. Investors already have high expectations for the company; slower growth or a gradual Guardian rollout could leave less room for error.

There’s a large market behind both pitches. Gartner forecasts worldwide information-security spending of about $244 billion in 2026, including demand for AI tools that strengthen defenses and products that protect companies as they use AI. The forecast describes a broad spending pool, though it doesn’t tell us how much either company will capture.

Good companies still need good trade setups

My process starts with the business, then turns to the market. I watch for unusual options activity (UOA), but treat it as a signal to investigate—not a guarantee that a stock will rise.

In August, a jump in options activity drew our attention to oil-services company SLB NV (SLB). I recommended a bullish trade. SLB shares rose about 9% over the following week, while the recommended trade gained 219%.

We also recommended a trade in MP Materials Corp. (MP) that returned 534% in three days. In both cases, the most a buyer could lose was the amount paid for each options contract. These were standout outcomes; options can also expire worthless. Position size and the amount at risk matter as much as spotting a promising setup.

When I see unusual activity, I check the news, the stock’s price and what I could lose. The next opportunity might be in cybersecurity, energy, health care or a company I hadn’t expected to follow that morning.

Putting the process to work

On October 14 at 8 p.m. EST, I’m hosting a free $10K to $100K Challenge. I’ll explain the strategy behind some of the trading opportunities I’ve highlighted over the years and how to apply that process.

Past recommended trades in BFLY and DOX gained 219% and 534%, respectively, after unusual options activity helped identify them. Earlier trades in OXY and BMY returned 779% and 1,076%. Those are individual past results, not promises of future performance. They show what can happen when a stock, a trade setup and risk management come together—but they don’t establish what a future trade will return.

You don’t need $10,000 to begin applying the process. I’ll discuss smaller trades, including opportunities where the amount at risk can be less than $500. Reserve your free spot in the $10K to $100K Challenge.

We’ll look at how unusual options activity can reveal where large investors are placing bets, then use that information to assess potential trades. There may be more opportunities as companies build out AI security, but the risks remain: a promising business doesn’t guarantee a successful trade, and a new product doesn’t guarantee customer demand.

Sign up for my free $10K to $100K Challenge.

Jonathan Rose
Founder, Masters in Trading

About the analyst

Jonathan Rose, editor of Masters in Trading: Daily Live, has traded and taught for more than two decades. He started on the Chicago Mercantile Exchange floor in 1997, managing Nasdaq and S&P 500 futures during the dot-com boom. Learn more about Jonathan.

Article printed from InvestorPlace Media: https://investorplace.com/dailylive/2026/10/openais-security-scare-points-to-2-cyber-stocks-worth-buying/. ©2026 InvestorPlace Media, LLC.

Source: Jonathan Rose · investorplace.com