🔥JPMorgan's list of "AI-Immune" software stocks is strong—but not complete

JPMorgan's list of AI-Resilient software stocks includes:

$Microsoft(MSFT.US)

$SentinelOne(S.US)entinelOne(S.US)erviceNow(NOW.US)

$Okta(OKTA.US)

$Twilio(TWLO.US)

$Palo Alto Networks(PANW.US)

$CrowdStrike(CRWD.US)

$Zscaler(ZS.US)

$SentinelOne(S.US)entinelOne(S.US)

$SentinelOne(S.US)entinelOne(S.US)NOW

$Datadog(DDOG.US)

$Veeva Systems -CL(VEEV.US)

$JFrog(FROG.US)

The core logic of this list is clear:

It's not about "who is best at AI,"

but rather—

who can maintain pricing power, stickiness, and cash flow stability in the AI era.

Roughly divided into three categories:

First category: Platform infrastructure

$Microsoft(MSFT.US)

$SentinelOne(S.US)entinelOne(S.US)erviceNow(NOW.US)

These companies directly embed AI into workflows,

customers won't churn because of AI,

but will rely more on the platform.

Second category: Security and identity

$Palo Alto Networks(PANW.US)

$CrowdStrike(CRWD.US)

$Zscaler(ZS.US)

$Okta(OKTA.US)

$SentinelOne(S.US)entinelOne(S.US)

AI expansion means an expanded attack surface.

Security budgets won't shrink, they'll only increase.

Third category: Data and observability

$SentinelOne(S.US)entinelOne(S.US)NOW

$Datadog(DDOG.US)

The higher the data density, the greater the value of AI.

But if you ask me—

what else can be added?

I would consider three supplementary lines.

First, the "computing efficiency layer":

$MongoDB(MDB.US)

Databases are the foundation of AI applications.

If AI applications explode, database read/write pressure will only increase.

Second, the "developer ecosystem control layer":

$GitLab(GTLB.US)

In the era of AI writing code,

whoever controls the development process controls the entry point.

Third, the "automation execution layer":

$UiPath(PATH.US)

For Agentic AI to be implemented,

process automation is needed as an execution framework.

AI will not eliminate software companies.

It will amplify companies that "own data, own customer entry points, and own security moats."

The real question is not—

who is doing AI?

But rather—

Who can still increase ARPU after AI becomes widespread, rather than being pressured on price?

Are you more inclined to bet on security,

or data platforms?

📬 I will continue to deconstruct the software layer structure of the AI era to identify companies with true shock-resistant capabilities. Welcome to subscribe and explore the direction of structural winners together.