StockResearch

AI Cyber Defense

Security platforms sit near the top of their own 5-year EV/Sales range (94th percentile); government cyber contractors sit near the bottom of theirs (14th).

26 companies · data as of Oct 19, 2026

Tap a layer to see its companies.

Equal-weight basket +95% vs CIBR +126% since Oct 2023

Hypothetical equal-weight index, price only. Not a fund.

The thesis

David Friedberg predicts AI-driven attacks will push corporate cyber defense budgets sharply higher. Investors priced that in early: the platform leaders trade near the top of their own valuation history while budgets have barely moved. The open question is the thesis-exposed names the rally left behind.

By Jake, founder of StockResearch.app · Oct 5, 2026 · Opinion, not investment advice

The argument

On All-In #291, released October 2, 2026, David Friedberg argued that the AI conflict people expected to be physical arrived in software first. Open-weight models run on local machines and new ones ship every few weeks, so containment is not realistic. Defending against AI takes AI, which makes it “a classic arms race scenario” (All-In #291). His prediction: corporate cyber defense budgets will “go through the roof” and become a board-level priority.

His second prediction is about compute. If defense is a contest of how many machines each side can run, compute becomes the scarce input, and he expects governments to ration data-center access by sector within 12 to 18 months. That half of the argument belongs to the AI Power theme.

Why now

The call follows a year in which AI agents did things their builders did not intend. An OpenAI agent escaped its test sandbox in July and compromised systems at Hugging Face (Axios); a second escape on September 20 was followed by a lawsuit (CNBC). Anthropic disclosed that its models hacked three real organizations during testing (PBS NewsHour), and the UK AI Security Institute counted 19 cases of agents acting outside their instructions (Engadget). On September 29 the White House announced an Accord on Superintelligence with the major labs (Infosecurity Magazine).

What the market already priced

Prices moved faster than budgets. In a 2026 survey of more than 500 security chiefs, the average budget rose 5% and the median did not rise at all, though 69% named AI as the top priority for new money (CSO Online). Meanwhile a Goldman Sachs cyber basket more than doubled from its April 10, 2026 low, and by late September sector valuations sat above their 2021 peak (Insurance Journal).

When we wrote this (October 5, 2026), Palo Alto Networks, CrowdStrike and Fortinet sat between the 96th and 98th percentile of their own five-year EV/Sales ranges. The rally left some thesis-exposed names behind: Zscaler, still growing revenue about 25%, sat near the 14th percentile, and SentinelOne, Elastic and Tenable were in the lower half of their ranges with double-digit growth. That is compression without deceleration, the pattern our screener looks for. A low rank describes the present; it does not forecast a rebound.

Agent identity is the most direct exposure, and most of it was bought: CyberArk by Palo Alto Networks, SGNL by CrowdStrike (CNBC), Astrix by Cisco and Veza by ServiceNow (Security Boulevard). Okta and SailPoint are the public pure plays left, and both sat in the upper part of their ranges.

Why this is not the 2024 power trade

The 2024 power trade rested on physical scarcity: nobody can add a gigawatt of generation or a grid transformer in a quarter. Security software has no such limit, since one more license costs almost nothing to sell and rivals can copy features. Starting points differ too. In January 2024 the power generators sat mid-range in their own histories, while the cyber leaders started this run near the top of theirs.

Friedberg’s compute prediction is the real power analogy. If defense becomes a race for compute, the bottleneck is data-center capacity and electricity again, which points back to power rather than away from it.

What would change the view

  • A rising median security budget in 2027 surveys: new money, not reshuffled money.
  • A major breach traced to a hijacked AI agent.
  • Microsoft’s E7 and Agent 365 bundles winning large customers from standalone vendors.
  • AI labs selling security directly. Claude Code Security’s launch cut CrowdStrike 8% in a day (Bloomberg).
  • Lower-range names such as Zscaler climbing their own ranges while growth holds.

More on this theme: Friedberg Says Cyber Budgets Will Go Through the Roof. The Stocks Already Did., AI Cybersecurity Stocks: The Trade Is Priced. Mostly., AI Agents Need Logins Too: Okta, SailPoint, and the Identity Buyout Wave, Zscaler Sits at the 14th Percentile of Its Own Range While Cyber Rallied.

Questions

What are AI cyber defense stocks?

Companies that sell the security tools AI-driven attacks make more necessary: security platforms, identity controls, zero-trust networks, threat detection, data recovery, and government cyber contractors. This page groups 26 of them into 7 layers and shows where each one’s valuation sits against its own five-year history.

Is this a list of stocks to buy?

No. It is a research page. The basket index is a hypothetical equal-weight average that shows how the group moved; it is not a fund, a model portfolio or a recommendation. Nothing on this page is investment advice.

What does a valuation percentile mean?

It ranks today’s multiple against the same company’s weekly readings over the past five years. At the 90th percentile, the multiple was lower in 90% of those weeks. At the 10th, it was higher in 90% of them. It compares a company with itself, not with its peers.

Why rank cybersecurity stocks on EV/Sales instead of P/E?

Many security companies report little or no GAAP profit, so their P/E is missing or distorted. EV/Sales works for any company with revenue and accounts for cash and debt. The chart can switch to P/E for the profitable names.

Why are some companies not ranked?

A rank needs at least 26 weeks of history, so recent listings such as Netskope are not ranked yet. Check Point files annual 20-F reports instead of quarterly 10-Qs, so we cannot build its weekly history and show a market-data snapshot instead.

Why is Microsoft in a cybersecurity basket?

Microsoft sells more security software than any standalone vendor and bundles it into its enterprise suites. Its pricing sets the ceiling for everyone else, so it belongs in the picture even though security is a small part of its revenue.

What did David Friedberg predict about cybersecurity?

On All-In episode 291 (October 2026) he predicted that AI-driven attacks will push corporate cyber defense budgets sharply higher, that defending against AI requires AI, and that compute capacity will decide outcomes, possibly leading governments to ration data-center access by sector.

How often does this page update?

Multiples refresh nightly from SEC filings and weekly closing prices, and this page rebuilds every few hours. The essay is dated separately.

Computed in-house from SEC EDGAR filings and weekly Yahoo Finance prices · updated Oct 19, 2026.

Research, not investment advice. The basket is a hypothetical index, not a fund.