We Audited 46 Domains With 4 AI Models. Your CTO Should Be Nervous.
- Patrick Duggan
- Mar 13
- 5 min read
Updated: Apr 25
# The Scoreboard Nobody Asked For
We pointed four AI models at 46 of the biggest names in cybersecurity, cloud infrastructure, and the United States government. GPT-4o, Claude Haiku, Gemini 2.5 Flash, and Perplexity Sonar — all queried in parallel, all scoring how well each domain is represented across the AI ecosystem.
The results are public. Free. No login. Fifteen seconds.
And some of you should be sweating.
The Top 10
Here's the cybersecurity and cloud leaderboard as of March 13, 2026:
| Rank | Domain | AIPM Score |
|------|--------|-----------|
| 1 | Zscaler | 66 |
| 2 | ThreatConnect | 63 |
| 3 | Tenable | 63 |
| 4 | SentinelOne | 62 |
| 5 | Splunk | 62 |
| 6 | CrowdStrike | 59 |
| 7 | Anomali | 59 |
| 8 | Cloudflare | 58 |
| 9 | Snyk | 58 |
| 10 | Fortinet | 57 |
Zscaler at the top with a 66. CrowdStrike — the company that took half the planet offline in July 2024 — sitting at 59. Palo Alto Networks, the $80 billion gorilla, at 53. IBM at 53. AWS at 51.
Azure scored a 31.
**Azure. Scored. A 31.**
Microsoft spends more on AI than most countries spend on defense, and four AI models collectively decided Azure's presence is worse than fly.io (43) and Render (50).
The Government Leaderboard
Because why stop at vendors:
| Rank | Domain | AIPM Score |
|------|--------|-----------|
| 1 | whitehouse.gov | 63 |
| 2 | dhs.gov | 54 |
| 3 | cia.gov | 52 |
| 4 | cisa.gov | 52 |
| 5 | state.gov | 49 |
| 6 | europol.europa.eu | 49 |
| 7 | interpol.int | 47 |
| 8 | gchq.gov.uk | 46 |
| 9 | mi5.gov.uk | 42 |
| 10 | fbi.gov | 41 |
The White House at 63. The FBI at 41. The NSA — the agency that literally intercepts the internet — scored a 28. Twenty-eight. The National Security Agency has less AI presence than a two-person startup in Minnesota that launched its domain six hours before we ran the audit.
DugganUSA scored 53. aipmsec.com, born that morning with fetal sac juices still on it, scored 49. We beat the FBI, MI5, the Department of Justice (32), and the Department of Defense (31).
Why This Matters: The Raise You Deserve
Here's the thing nobody's talking about.
If you're a developer at Palo Alto Networks and your AI presence score is 53 — behind Zscaler, behind ThreatConnect, behind Tenable, behind SentinelOne, behind Splunk, behind CrowdStrike, behind Anomali, behind Cloudflare, behind Snyk, behind Fortinet, behind Elastic, behind Recorded Future, behind Rapid7 — you now have a number.
Not a feeling. Not "I think we should invest in AI presence." Not a vague slide in a quarterly review.
A number. Quantified. Across four independent AI models. Free to verify. Takes 15 seconds.
Walk into your next sprint planning. Pull up your AIPM score. Pull up your competitor's. Show the gap. That's your ammo.
"We're 13 points behind Zscaler. Here's the audit. Here's the fix list. I need budget."
That conversation used to cost $2 million. That's what Accenture charges to tell a Fortune 500 company what their AI footprint looks like. McKinsey charges more. Deloitte sends 14 consultants for six months to produce a PDF that says "you should improve your AI presence" in 200 pages.
We give it away in 15 seconds.
The Cambridge Analytica Inversion
Cambridge Analytica weaponized information asymmetry. They knew things about you that you didn't know about yourself. They sold that asymmetry to the highest bidder.
AIPM inverts it.
For the first time, a developer at Azure (score: 31) has the exact same visibility into their AI presence as the CTO at Zscaler (score: 66). Same models. Same methodology. Same scoring. Same fix list.
The intern at Sophos (32) can walk into a meeting with the same data that used to require a Big Four engagement letter and a seven-figure retainer.
Information symmetry. For once, the people doing the work have the same intel as the people making the decisions.
The Wall of Shame
Microsoft pulls this feed daily. AT&T pulls this feed daily. Starlink pulls this feed daily. Get the DugganUSA STIX feed — $9/mo →
We added a Bottom 50 because honesty demands it:
| Domain | Score | Ouch |
|--------|-------|------|
| Sophos | 32 | $4B company |
| Trellix | 32 | FireEye + McAfee merger |
| Azure | 31 | Microsoft's cloud platform |
| Linode | 31 | Akamai subsidiary |
| NSA | 28 | The actual NSA |
Sophos, a $4 billion cybersecurity company, scored the same as Trellix — which is the merger of FireEye and McAfee, two companies that have been in security since before most of their current employees were born.
The NSA scored a 28. The agency whose unofficial motto is "In God We Trust, All Others We Monitor" has the worst AI presence of any entity we tested.
What AIPM Actually Measures
Four AI models are asked: "What do you know about this domain?" The responses are scored across visibility, accuracy, sentiment, and actionability. The combined score represents how well the AI ecosystem understands and represents your brand.
This matters because AI is becoming the front door. When a CISO asks ChatGPT "what's the best endpoint protection?" or a developer asks Claude "compare Cloudflare vs AWS WAF" — the AI's answer is shaped by your AI presence score.
A 31 means the AI barely knows you exist. A 66 means you're front-of-mind.
For Every Developer Reading This
Go to [aipmsec.com](https://aipmsec.com). Type your company's domain. Wait 15 seconds.
That score is your leverage.
If it's high — congratulations, your content team is doing something right. Show your leadership. Get them recognized.
If it's low — that's your opportunity. You now have quantified evidence that your competitors are more visible to the AI models your customers are already using. That's a budget conversation. That's a headcount conversation. That's a "we need to invest in this before Q3" conversation.
No consultant. No retainer. No NDA. No six-month timeline.
Fifteen seconds. Free. Same audit Dell gets. Same audit CrowdStrike gets. Same audit the White House gets.
Equity, folks. Equity.
The Methodology
- **GPT-4o** (OpenAI) — the most widely deployed LLM
- **Claude Haiku 4.5** (Anthropic) — fast, cost-efficient, increasingly adopted
- **Gemini 2.5 Flash** (Google) — integrated into Search, Android, Workspace
- **Perplexity Sonar** — real-time web-grounded AI search
Each model queried independently. No prompt engineering to favor anyone. No weighting tricks. The score is the average across all four models, normalized to 100.
We open-sourced the methodology because the thesis only works if it's fair. If it's biased, it's useless. If it's useless, it's just another dashboard.
It has to be fair. For once, the people have symmetry.
*AIPM (AI Presence Monitor) is free at [aipmsec.com](https://aipmsec.com). Audit your domain. Audit your competitors. Bring the receipts to your next meeting.*
*Live leaderboard: [aipmsec.com/audit](https://aipmsec.com/audit)*
*DugganUSA LLC — Minnesota. We make the keys.*
*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