PLTR: The Only Software Company The Pentagon Cannot Replace

Palantir is the single most polarizing stock in the US equity market. The bears think it is a glorified consulting business with a software wrapper, overvalued by a factor of three, supported by a cult-like retail base. The bulls think it is the most important applied AI software company on the planet, the only true operating system for enterprise and government AI deployment at scale.

I have been on both sides of this debate at different points. After watching the company execute through 2024 and 2025, after watching the AIP commercial revenue growth, after seeing the government contract pipeline expand under multiple administrations, after working through the margin profile, the cash flow conversion, and the customer base, I am firmly in the bull camp.

Here is the full case, the risks, and the way I think about owning it.

Palantir owns the most defensible enterprise AI operating system in the world, with a government franchise that the US national security apparatus cannot meaningfully replace, and a commercial business that is just beginning to compound at scale.

Palantir sells three core platforms. Gotham is the government platform used for intelligence, defense, and law enforcement workflows. Foundry is the commercial platform used by enterprises to integrate data and build operational software. AIP is the artificial intelligence platform that sits on top of both and lets customers build, deploy, and govern AI agents on their own data inside their own environments.

The simple version is this. Palantir takes a customer's messy data, integrates it, makes it usable, and then lets the customer apply AI to it to make real operating decisions in real time. Inventory routing. Drone targeting. Insurance claims processing. Hospital workflow optimization. Manufacturing line balancing. Whatever the operational decision is, AIP wants to be the platform on which it happens.

This is the part the bears underestimate. Palantir's government work, particularly with US defense and intelligence customers, is built on over a decade of accumulated security clearances, accredited environments, embedded forward-deployed engineers, and operational integration across the most sensitive workflows the federal government runs.

Replacing Palantir in those workflows is not a procurement question. It is an institutional capability question. The contracts get bigger over time, not smaller, because every new application built on the platform increases the cost of moving off it. The Army's TITAN program. The Marine Corps. Special Operations. The intelligence community. The Department of Defense Maven AI program. These are not one-off contracts. They are platform commitments.

The DOD CIO has publicly signaled that platform-based software acquisition is the future. Palantir is the incumbent. The political winds across both parties favor accelerating defense software modernization. The contract pipeline keeps getting larger.

For years the bear case on Palantir was that government revenue was the only meaningful business and the commercial side was struggling to break out of pilots. AIP changed that. The release of AIP in 2023 and its rapid commercial adoption through 2024 and 2025 produced the inflection the company had been promising for years.

Commercial revenue growth accelerated. Customer count expanded. The AIP boot camp model proved out as a faster, cheaper, more repeatable customer acquisition motion. Companies that had spent five years debating whether Palantir's pricing was justified now sign in 90 days because the AIP value is operational and measurable.

This is the part of the story that took the longest to play out and is now actually playing out.

Driver one: AIP commercial flywheel. Every quarter commercial customer growth and dollar-based net retention build on the prior quarter. Each AIP win produces case studies that accelerate the next AIP win. The motion is now repeatable.

Driver two: government tailwind. US defense and intelligence software modernization is accelerating. Allied governments are following. Palantir is the platform of choice for the largest workflows. Contract sizes and durations keep extending.

Driver three: profitability inflection. Palantir has crossed into GAAP profitability and is generating meaningful free cash flow. Rule of 40 metrics are improving. The margin structure looks more like a software platform every quarter.

Driver four: index inclusion and institutional adoption. S&P 500 and Nasdaq 100 inclusion have brought passive flows into the stock. Institutional ownership is rising as long-only funds finally get comfortable with the model. That structural buyer base reduces volatility versus the meme-stock era.

US commercial revenue growth. Government revenue growth. Total customer count. Dollar-based net retention. Operating margin expansion. Free cash flow conversion. AIP boot camp count and conversion rate. Remaining performance obligations.

The single metric I watch most closely is US commercial growth. If that line keeps compounding, the bull case is intact.

Palantir trades at a premium revenue multiple. That is not new. It has traded at a premium multiple for most of its public history. The premium is justified by the moat, the growth rate, the operating leverage, and the strategic position. The premium is also a risk if growth ever decelerates.

I do not own Palantir for multiple expansion from here. I own it because I believe the operating story compounds for many years and that earnings growth eventually catches the multiple.

The first risk is valuation. The stock can compress sharply on any soft quarter. That is the cost of owning a premium multiple name.

The second risk is execution. AIP boot camp conversion is the engine right now. Any meaningful slowdown is a real concern.

The third risk is political. Government contracts are political. A change in administration view on Palantir's customer base could pressure specific contracts even though the overall federal software trend is favorable.

The fourth risk is competition. Microsoft, Databricks, Snowflake, and ServiceNow are all moving into adjacent enterprise AI territory. Palantir's moat is real but is being attacked from multiple directions.

The fifth risk is dilution. Stock-based compensation has been high. That is improving but remains a watch item.

Palantir is a core long-term position with a higher volatility profile than my software comp set. The right size is meaningful, not maximal. I add on real selloffs driven by sentiment rather than fundamentals. I do not trim into strength unless valuation gets truly absurd.

There is no comparable company. Nobody else has the government depth, the commercial AI platform, the operational deployment muscle, and the profitability profile combined. You can argue with the multiple. You cannot argue with the strategic position.

That is why I own it.

Watch a recent AIP boot camp recording or customer case study on Palantir's investor site. Read the most recent earnings transcript. Look at the US commercial revenue growth line specifically. Compare to what the company guided to three quarters ago. Then ask whether you think the company is executing. If yes, you understand why I own it. If no, this is a name to revisit later.

Read next: NVDA: Why The King Is Not Done | ARM: The Royalty Machine

*Disclaimer: MentorSurge is not a financial advisor and this is not financial advice. This post is for educational and entertainment purposes only. Nothing here is a recommendation to buy or sell any security. Palantir is a high-multiple growth stock subject to significant volatility. Investing involves substantial risk of loss. Numbers cited were accurate when written and change constantly. Always do your own research and consult a licensed professional before making decisions with real money.*