AI Rewired Cybersecurity: The Patch Race and Who Gets Paid

AI has changed the cybersecurity business in three ways at once: it finds software flaws faster than companies can fix them, it opens new ways into AI agents, and it is pushing customers toward fewer, larger security platforms.

Our number: Palo Alto Networks agreed to pay about $4.5 billion in stock for Applied Compute, equal to about 46% of its own $9.73 billion of Next-Generation Security annual recurring revenue, by our arithmetic from its results and deal release.

Business Pill · Least Privilege

An evergreen Business Engineer explainer on least privilege and access control. It teaches the general idea only and does not discuss Palo Alto Networks or this analysis.

The key insight: AI speeds up both finding and fixing software flaws, makes AI agents a new target, and is moving security spending to platforms: Palo Alto’s NGS ARR grew 66% against 39% for revenue, and it agreed to pay stock worth about 46% of that ARR for Applied Compute, by our arithmetic.

The Patch Race

The first change is speed. AI tools now find exploitable flaws quickly, and fixing them takes longer than finding them.

Palmer Luckey put it this way at the New York Times DealBook Summit, TechCrunch reports: “We’re seeing the natural effects of a lot of these advanced AI tools allowing for the creation of various exploits.”

Microsoft’s Satya Nadella pointed at what cannot be fixed at all. “Remember, there is a lot of stuff in our critical infrastructure that’s unpatchable,” he said, per TechCrunch, naming firmware, chillers and HVAC systems.

The patching itself can break things. Sources told The Washington Sun that the NSA and bank executives briefed Congress that fixes for AI-found flaws drove recent outages.

Officials did not call those outages attacks. Director of National Intelligence Jay Clayton told a House hearing they were “separate service outages associated with maintenance and technical changes”, CBS News reports.

Palo Alto Networks' Next-Generation Security ARR grew 66% year over year to $9.73 billion in fiscal Q1 2027, a
Palo Alto Networks’ Next-Generation Security ARR grew 66% year over year to $9.73 billion in fiscal Q1 2027, against 39% for total revenue and 35% for remaining performance obligations. Source: Palo Alto Networks results release.

What Was Posted on X

Wall St Engine posted Palo Alto Networks’ fiscal Q1 2027 highlights on 18 November 2026, including revenue of $3.44 billion, Next-Generation Security ARR of $9.73 billion and the raised full-year guide. This analysis takes those figures from the company’s own release.

The Information posted on 18 November 2026 that its reporter spoke with OpenAI’s Greg Brockman and CFO Sarah Friar, who said cybersecurity could be a bigger opportunity for the company than coding. This analysis cites that report in the section on the AI labs.

Update, 19 November 2026, 00:41 UTC: UK financial firms are told to tighten cyber defences as AI exposes vulnerabilities, the FT reports. The Financial Times ran the headline “City firms told to tighten cyber defences as AI exposes vulnerabilities” (FT); we could read only the headline and its X post, not the article. The UK authorities set out their expectations earlier. In a joint statement published on 15 May 2026, the Bank of England, the FCA and HM Treasury said “The Government and UK financial authorities judge that firms should be taking active steps across several domains” and “It is essential that firms have effective protective, detective, threat containment and cyber response capabilities” (Bank of England). The statement also told firms to “consider adopting automated and AI-enabled defences to operate at comparable speed to AI-driven attacks”. Related: Palo Alto lifts its revenue guide and buys Applied Compute.

The New Attack Surface: Agents

The second change is that AI agents themselves are targets. Hidden instructions in emails and documents steered agents in 34.9% of PromptArmor’s FrontierBreach tests across 21 models.

PromptArmor sells AI-security products, so that is a vendor benchmark. It still shows why buyers now ask security vendors to cover agents, not only networks and laptops.

Consolidation: Fewer, Bigger Platforms

The third change is who gets paid. Palo Alto’s results show money moving to platform subscriptions faster than to the company overall.

Its Next-Generation Security ARR grew 66% from a year earlier to $9.73 billion, while total revenue grew 39% to $3.44 billion, the release says. Remaining performance obligations rose 35% to $21.0 billion.

Management tied the outlook to the threat level. “Given the current threat environment and demand we see ahead of us, we are materially raising our full year guidance,” said finance chief Dipak Golechha.

The fiscal 2027 revenue guide rose by about $395 million at the midpoint, and the NGS ARR guide by about $375 million, by our arithmetic against the outlook it gave on 1 September.

Buying the AI Layer

Applied Compute helps companies train, adapt and deploy open-weight AI models. Chief executive Nikesh Arora said the deal will let Palo Alto deliver “secure open weight” AI to enterprises.

The 8-K values the deal’s shares at $337.16 each. At that price $4.5 billion is about 13.3 million new shares, roughly 1.6% of shares outstanding, by our arithmetic; the 8-K says the final count is not yet known.

Investors had bid the whole group up before the results. Eight cybersecurity stocks rose an average of 8.64% on 18 November, by our arithmetic from delayed quotes.

The AI Labs Want In

The model makers see the same market. OpenAI’s Greg Brockman and finance chief Sarah Friar told The Information that cybersecurity could be a bigger opportunity than coding for the company.

That sets up a question for the incumbents: whether security stays a separate product category or becomes a feature that frontier labs sell alongside their models.

The Structural Read

As we read Palo Alto’s release, the company is linking its guidance directly to the threat level, not only to product demand.

As we read the deal terms, Palo Alto has agreed to pay for an AI-model capability with stock rather than cash, about 1.6% of its shares by our arithmetic.

Dipak Golechha, chief financial officer, Palo Alto Networks, in the Q1 FY27 results release

“Given the current threat environment and demand we see ahead of us, we are materially raising our full year guidance.”

Three Implications

FOR SECURITY BUYERS Agents add a new surface: FrontierBreach found hidden instructions steered agents in 34.9% of tests, a vendor benchmark.

FOR INCUMBENTS Palo Alto’s release shows platform subscriptions (NGS ARR +66%) outgrowing revenue (+39%).

FOR OPERATORS Nadella, per TechCrunch, said much critical-infrastructure firmware is unpatchable and connected to the internet.

The Business Engineer Lens

This story maps onto the Business Engineer framework The Open vs Closed Meta-Framework.

The framework starts from this line: “Every technology stack has layers. Some are scarce, some are abundant.”

As we read the Applied Compute deal, Palo Alto is buying a way to secure open-weight models, the layer the framework treats as abundant, rather than building its own frontier model. The framework predates the deal and does not discuss it.

What Is Not Established

No source here measures how much security spending AI has added overall. Palo Alto’s figures cover one company, and the outage explanation comes from unnamed sources, not the organisations involved.

This Week In The AI Supercycle

AI Lost Its Central Banker, the Lab Oligopoly, and the Future of Open Source

This week’s issue: the Nvidia leadership change, read as the AI build-out losing its central banker: who now finances the GPU landlords, backstops the data centres and sponsors open models, and why two labs are becoming the bank.

Read the full issue →

Business Engineer Framework

The Open vs Closed Meta-Framework

A Business Engineer framework on which layers of a technology stack to keep proprietary and which to open.

Read the Map of AI →

The Bottom Line

AI speeds up both attack and repair, and opens agents as a new target. Palo Alto’s NGS ARR grew 66% against 39% for revenue, and it agreed to pay stock worth about 46% of that ARR for an AI-model company, by our arithmetic.

97,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.

A note on sourcing. Palo Alto Networks’ figures and executive quotes come from its 18 November 2026 results release, its Applied Compute release and its Form 8-K. Luckey’s and Nadella’s remarks are from TechCrunch’s DealBook reports; Clayton’s is from CBS News; we did not watch the sessions. The NSA and banks briefing account and the FrontierBreach figures are from our linked earlier pieces and their sources. The 46%, $395 million, $375 million, 13.3 million share and 1.6% figures are our arithmetic. Not investment advice; for information only.

Sources: Bank of England, FCA and HM Treasury joint statement on frontier AI and cyber resilience (15 May 2026) · Palo Alto Networks: Fiscal First Quarter 2027 results · Palo Alto Networks: intent to acquire Applied Compute · CBS News: officials find no evidence outages were cyberattacks · TechCrunch: Nadella says critical infrastructure is unpatchable · TechCrunch: Palmer Luckey warns AI-powered cyberattacks are outpacing defenses

Scroll to Top
FourWeekMBA