Nikesh Arora sat down and purchased $10 million worth of shares in his own company on a Friday afternoon in late March, as cybersecurity stocks were plummeting across the board and trading desks were experiencing yet another wave of anxiety about what artificial intelligence might do to the industry. Not via an automated trading plan that has been prearranged.
Not by giving executives stock-based compensation, which they may or may not accumulate. An open market purchase is the type in which you determine that the stock is less expensive than it should be and that you want more of it, fully aware of everything the market is pricing in. He hadn’t made a purchase like that since November 2019. Almost as important as the number is that gap.
| Person | Nikesh Arora |
|---|---|
| Title | Chief Executive Officer, Palo Alto Networks |
| Previous Roles | President & COO, SoftBank Corp.; Senior Vice President & Chief Business Officer, Google |
| Company | Palo Alto Networks, Inc. (Nasdaq: PANW) |
| Headquarters | Santa Clara, California, USA |
| Transaction Date | March 28, 2026 (disclosed via SEC filing) |
| Shares Purchased | 68,085 shares |
| Total Purchase Value | ~$10 million (at $146.46–$147.48 per share) |
| Last Previous Open Market Buy | November 2019 |
| Arora’s Total PANW Holdings Post-Purchase | ~$162 million |
| PANW Stock Price at Time of Purchase | ~$147 |
| PANW 52-Week High | $224 |
| PANW YTD Performance (2026) | −14% |
| Most Recent Quarterly Revenue Growth | 15% year-over-year |
| Next-Gen Security ARR | $6.33 billion (+28% YoY) |
| Reference Website | Palo Alto Networks Investor Relations |
The amount was not symbolic, and the timing was not coincidental. Before that Friday session, Palo Alto Networks had been steadily losing ground in 2026, down about 14% for the year. This was due to a narrative that had been developing for months throughout the cybersecurity industry. If artificial intelligence can automate the vulnerability scanning, threat detection, and incident response that businesses like Palo Alto charge tens of thousands of dollars annually for, then what exactly are investors paying for?
This is a simple fear that, if taken at face value, is truly unsettling for anyone who owns these stocks. When Anthropic unveiled a tool in February that could scan code for security flaws—exactly the kind of thing that has historically required specialized security software and the companies that build it—the concern grew. Days prior to Arora’s acquisition, there was a rumor that Anthropic was working on a more potent model that might make it easier for hackers to launch attacks. Stocks in cybersecurity fell once more. Arora made a purchase.
In one version of this tale, $10 million hardly makes an impression. As CEO, Arora makes about $100 million a year, and after the acquisition, his total ownership stake in Palo Alto Networks is about $162 million. Skeptics might legitimately contend that $10 million is a courteous gesture from someone for whom $10 million is a rounding error. A YouTube financial commentator made precisely this point just hours after the filing was made public.
That interpretation overlooks a crucial aspect of how insider purchases serve as signals in reality. The exact amount in relation to Arora’s net worth is not the pertinent question. It’s the choice to make an open market purchase at all, following a six-year hiatus, in an industry that is being actively attacked ideologically by a narrative that raises concerns about whether cybersecurity software as a whole is becoming outdated. That’s a specific, intentional decision, and executives don’t make it lightly because if the stock continues to decline, they will appear foolish in public and suffer personally.
Alongside the acquisition, Arora wrote a blog post the next Monday in which she argued that cybersecurity firms and AI labs should collaborate rather than compete. He referred to this as the “most consequential” time in the industry’s history and urged swift, concerted action on both sides.
The CEO of one of the biggest cybersecurity firms in the world is rejecting the idea that artificial intelligence poses a threat to his company, despite the cautious wording. He contends that rather than reducing the need for defense, AI increases the surface area of possible attacks, a claim that is also present in Palo Alto’s most recent product releases. Increased entry points, complexity, and potential for exploitation result from more AI agents functioning independently across enterprise systems. According to that perspective, the development of AI does not pose a threat to cybersecurity. It is a driver of demand.
Even though the stock’s 14% decline indicates that the market is still not convinced, the product story Palo Alto is developing around that thesis is worth taking seriously. The company recently released Prisma AIRS 3.0, which was created especially to secure autonomous AI agents, the type of software entities that are currently being deployed across corporate networks by Microsoft, Google, and numerous enterprise software vendors.
Additionally, it unveiled Prisma Browser for Business, which targets smaller businesses that are increasingly using AI-heavy workflows, and Next-Generation Trust Security, which aims to automate certificate management in a post-quantum encryption world. In the meantime, the company closed two large acquisitions: Chronosphere, an AI observability platform, and CyberArk, an Israeli identity security company. A total of $4.9 billion in cash was spent on the AI security opportunity that Arora continues to publicly and now personally highlight. The company reported 15% revenue growth and operating margins above 30% for the third straight quarter. The company is not in danger. At the moment, the stock price simply does not reflect that.
It’s difficult to ignore the similarities to previous instances in which industry-wide narratives pushed stocks below levels where company fundamentals supported the price. In 2022, the market began to question whether high-multiple software companies were worth their valuations due to rising interest rates, which caused cloud software stocks to undergo a similar reckoning.
Eventually, the businesses with growing margins, real revenue growth, and strong product positions made a comeback. Without those foundations, those who had been riding the wave did not. It seems that it will take some time for the differences between the two groups to become apparent as the cybersecurity industry absorbs the AI disruption thesis in early 2026. Arora is betting Palo Alto falls firmly into the first category.
At least some validation was provided by the market’s initial response. Following the announcement of the acquisition, Palo Alto’s stock increased by about 5% on Monday, while CrowdStrike, Okta, and Netskope all saw gains of about 3% in sympathy. This suggests that Arora’s personal capital commitment changed sentiment not only for one company but for the industry as a whole.
The insider purchase was referred to by JPMorgan as a “substantial vote of confidence,” a term used by analysts to imply agreement without explicitly stating it. As the deeper concerns regarding AI’s impact on cybersecurity spending manifest in real enterprise purchasing decisions over the coming quarters, it’s still unclear whether that sentiment holds. Those questions are not addressed by a $10 million stock purchase. However, it does provide you with a clear and costly indication of the direction that the most knowledgeable person believes the answer will take.
