Palo Alto Networks Sells Visibility. Their Website Is Invisible to AI.
- Patrick Duggan
- Mar 12
- 6 min read
Updated: 2d
# Palo Alto Networks Sells Visibility. Their Website Is Invisible to AI.
**Published:** March 12, 2026
**Author:** Patrick Duggan, DugganUSA LLC
**Category:** AIPM / Analysis / Cybersecurity
I used to work there.
Not in a "I was an intern" way. I was embedded at Microsoft for the JEDI contract, came through Dell EMC, and spent real time inside the Palo Alto Networks machine. I know what the building looks like from the inside. I know what the culture smells like. I know what ships and what doesn't.
So when I pointed MedusAIPM at paloaltonetworks.com and the structure score came back **25 out of 95**, I didn't laugh. I sighed.
Because of course it did.
The Audit
MedusAIPM scores companies on AI Perception (how GPT-4o, Gemini, and Perplexity describe your brand) and Structural Readiness (whether AI crawlers can machine-read your website). Combined score out of 95.
| Metric | Score |
|--------|-------|
| **Combined AIPM** | **53/95** |
| **AI Perception** | **72/95** |
| **Structure** | **25/95** |
| **AIPM-NPS** | **0** |
Twenty-five. A $100 billion cybersecurity company — the one that invented the next-generation firewall, the one with 16,000 employees and $9.2 billion in revenue — scores 25 on structural AI readiness.
For context, here's the cybersecurity leaderboard:
| Company | Revenue | Combined | Structure |
|---------|---------|----------|-----------|
| **Zscaler** | $2.3B | **72/95** | **76/95** |
| **CrowdStrike** | $3.9B | **69/95** | **67/95** |
| **DugganUSA** | Pre-revenue | **66/95** | **86/95** |
| **Palo Alto Networks** | $9.2B | **53/95** | **25/95** |
The largest cybersecurity company in the world, by revenue, has the worst AI presence in the industry. A pre-revenue startup from Minnesota outscores them by 61 points on structure.
What Zero Looks Like
Let's walk through the structure scores:
- **robots.txt:** 95/95. They explicitly allow GPTBot, ClaudeBot, PerplexityBot, and every major AI crawler. Someone on the infrastructure team did this correctly. The front door is open.
- **LD-JSON:** 0. Zero structured data. No Organization schema. No Product schema. No FAQ schema. Nothing. The AI crawlers walk through the open door and find an empty room.
- **Semantic HTML:** 0. No `<article>`, no `<section>`, no `<main>`, no `<nav>`. Just divs. An ocean of divs. The page renders beautifully for humans and tells AI models absolutely nothing about what it contains.
- **Sitemap:** 10. Barely functional.
- **Meta tags:** 0. No OpenGraph. No Twitter Card. No meta descriptions that crawlers can parse.
Read that sequence again. They opened the door to AI crawlers — explicitly, in robots.txt — and then gave them nothing to read. That's worse than blocking crawlers entirely. At least blocking is a decision. This is negligence.
The Irony Is Structural
Palo Alto Networks sells **visibility**. That's the entire pitch.
- **Cortex XDR**: "See everything. Miss nothing."
- **Prisma Cloud**: "Complete visibility across your cloud estate."
- **Unit 42**: Threat intelligence based on analyzing 30 million malware samples daily.
They sell the ability to see what others can't. Their entire $100 billion valuation is built on the premise that they can make the invisible visible.
Their website is invisible to AI models.
Not because they lack brand awareness — GPT-4o knows exactly who they are (85/95 awareness). Perplexity wrote a 350-word essay about their product lines, revenue, Nir Zuk's founding story, and their Gartner Magic Quadrant positions. The models *know* Palo Alto Networks.
But that knowledge comes from training data — a snapshot that goes stale. When GPT-5 and Gemini 3 re-crawl the web to update their understanding, paloaltonetworks.com will hand them nothing. No structured data describing their products. No schema explaining their services. No machine-readable identity.
The company that sells visibility will become whatever the training data says they were. Past tense.
The Competitors Did the Work
Zscaler — smaller company, less revenue — has a structure score of 76/95. LD-JSON at 85. Semantic HTML at 75. Meta tags at 95. Someone at Zscaler understood that AI crawlers are the new search engines and built accordingly.
CrowdStrike: 67/95 structure. LD-JSON at 65. Meta tags at 95.
Even DugganUSA — two people, zero revenue, $500/month infrastructure — has LD-JSON at 85, meta tags at 95, and a structure score of 86. We implemented Schema.org structured data in an afternoon. Eight structured data blocks on our main page. Organization schema, FAQPage schema, WebApplication schema, Product schemas.
Palo Alto Networks has 16,000 employees and zero structured data blocks.
This isn't a resource problem. Zscaler did it. CrowdStrike did it. We did it. The resources exist inside Palo Alto Networks — they have more web developers than our entire company has employees. This is a priorities problem.
What AI Models Actually Say
Microsoft pulls this feed daily. AT&T pulls this feed daily. Starlink pulls this feed daily. Get the DugganUSA STIX feed — $9/mo →
GPT-4o gave Palo Alto Networks an **8 out of 10** confidence rating. Solid. Expected for a market leader.
Perplexity gave **10 out of 10**, with an encyclopedic answer citing $9.2B revenue, 16,000 employees, Gartner leadership positions, and Nir Zuk's founding story.
Gemini gave them a **6 out of 10** and refused to recommend: *"As an AI, I do not have personal opinions or the ability to recommend companies."*
AIPM-NPS: **0**. One promoter (Perplexity), one passive (GPT), one detractor (Gemini). Net zero. For context, CrowdStrike's NPS is 67.
The models that have strong training data on Palo Alto rate them well. The models that rely more on real-time crawling — or that are just more cautious — rate them lower. That's the gap between awareness and readiness. Awareness decays. Readiness compounds.
I Know Why This Happened
I worked there. I know the culture.
Palo Alto Networks is an engineering-led company that measures itself by threat prevention efficacy, product integrations, and customer renewals. The website is a marketing function. Marketing doesn't ship firewalls. In the internal hierarchy, the people who build Cortex XDR outrank the people who build paloaltonetworks.com by three levels of organizational gravity.
Nobody at Palo Alto Networks wakes up thinking about Schema.org markup. They wake up thinking about zero-day exploits, cloud misconfigurations, and next quarter's ARR.
That's understandable. It's also how you end up with a $100 billion company that has a worse AI presence than a pre-revenue startup from Saint Paul.
The fix is the same as it always is: one developer, one afternoon, five Schema.org blocks. Organization, Product, FAQ, BreadcrumbList, WebSite. Add OpenGraph tags. Add meta descriptions. Ship it.
The question isn't whether they can. It's whether anyone in the building cares enough to file the ticket.
The Bigger Problem
Palo Alto Networks isn't special here. The entire cybersecurity industry — an industry that literally sells digital awareness — is behind on AI presence.
| Company | Structure Score |
|---------|----------------|
| Zscaler | 76 |
| CrowdStrike | 67 |
| DugganUSA | 86 |
| Palo Alto | 25 |
The company that coined "next-generation firewall" is running a last-generation website. The company that tells customers to adopt Zero Trust has zero structured data. The company that sells Cortex — named after the part of the brain that processes information — has a website that AI brains can't process.
I'm not angry. I'm just not surprised.
Audit Your Own Domain
MedusAIPM is in enterprise beta. If you want to know how AI models see your brand:
**API:** `POST analytics.dugganusa.com/api/v1/aipm/audit`
**Free tier:** [epstein.dugganusa.com/pricing](https://epstein.dugganusa.com/pricing)
The company that sells you visibility might want to check their own.
*MedusAIPM is Patent #102 in the DugganUSA portfolio. Part of the MEDUSA product suite.*
*DugganUSA LLC | Saint Paul, Minnesota | dugganusa.com*
**Methodology:** All audits conducted March 12, 2026 using MedusAIPM v1.0.0. AI perception scores from GPT-4o, Gemini 2.5 Flash, and Perplexity Sonar. Structure scores from automated homepage crawling. Combined = 60% perception + 40% structure. Epistemic cap: 95/95.
**Disclosure:** The author is a former Palo Alto Networks affiliate. This audit uses the same methodology applied to all companies in the MedusAIPM leaderboard, including DugganUSA.
*Her name was Renee Nicole Good.*
*His name was Alex Jeffery Pretti.*
The cheapest, fastest, most accurate threat feed on the internet.
275+ enterprises pulling daily. 1M+ IOCs. 17.4M indexed documents. We beat Zscaler by 43 days on NrodeCodeRAT. Starter tier $9/mo — less than any competitor’s sales demo.
Comments