When OpenAI and Anthropic enter the market to "steal business", Palantir's real competitor has arrived.

Palantir (PLTR.US) is currently standing at a delicate crossroads: It has undergone a transformation with the help of the AI wave, but it may also risk being engulfed by the same wave. According to Wall Street News, after the US stock market closed on Monday, the 4th of this month, Palantir announced impressive financial results, breaking historical records in both revenue and profit, with a year-on-year sales increase of over double in the US market. However, the company's stock price has fallen nearly 20% year-to-date, deviating significantly from its fundamentals. Wall Street is betting that with the rapid expansion of large AI labs such as OpenAI and Anthropic, Palantir's core software product's appeal to customers may gradually be eroded. According to The Wall Street Journal, OpenAI is building a data connection and structuring platform that is believed to directly compete with Palantir, with former Palantir employees also on the team. Both OpenAI and Anthropic have also replicated Palantir's iconic "forward deployment engineer" model, deploying engineers embedded in customer teams to drive AI implementation. William Blair analyst Louie DiPalma straightforwardly stated in a recent research report: The competition from Anthropic and OpenAI is intensifying for Palantir. AI is the engine, but also the hidden danger Palantir's business model is built on data integration and analysis. Palantir helps government agencies and corporate clients extract insights from massive amounts of information, supporting scenarios such as supply chain planning and military decision-making. The company only achieved profitability for the first time in 2023, over twenty years after its founding. The rise of AI has opened a new window for Palantir: enterprise customers are increasingly using OpenAI, Anthropic, and Alphabet Inc. Class C's large language models within the Palantir data platform, directly driving the explosive growth of its revenue. Since Palantir launched its Artificial Intelligence Platform (AIP), its stock price has surged over 2300%. (Palantir officially launched its Artificial Intelligence Platform in April 2023, followed by continued stock price strength). However, Palantir is not fundamentally an AI company. It does not develop its own models, but the operational logic of AIP is to introduce third-party models to enhance its own software capabilities. To use the words of Palantir employees, large language models are like crude oil, and Palantir is the refinery that refines it into a usable product. But more and more people are starting to believe that this "crude oil" will eventually learn to refine itself. Some experts estimate that large language models are already capable of reproducing most of the work Palantir has done in large-scale data understanding. Jake Behan, managing director of investment firm Direxion Capital Markets, succinctly summed up the controversy: The debate around Palantir is not about growth, but about whether it is an indispensable part of the AI technology stack, or just an expensive package wrapped around increasingly cheap AI models. Executives talk tough, but data speaks Faced with external scrutiny, Palantir's management displayed a consistent tough attitude at this week's investor conference call. The executives referred to the output of large AI labs as "slop," a term that was repeatedly used 17 times throughout the call. CEO Alex Karp stated: Enterprise customers can try various AI products on the market, but most will eventually come back to Palantir. CTO Shyam Sankar stated that cheaper open-source models have actually brought more business to Palantir. He said: As models become better, cheaper, and more capable, we benefit more. These labs are not our competitors; they are our supply chain. However, cracks have already appeared in the financial data. The growth rate of Palantir's US commercial contract bookings dropped from 137% in the previous quarter to 45%. This discrepancy has raised analyst concerns, suggesting that the momentum of commercial market expansion is weakening, and the transmission effect of competitive pressure may already be showing. Government moat, deep but not endless On the government side, Palantir's barriers remain strong. With its early advantage in the defense sector and deep political resources in Washington, Palantir won over $1.1 billion in federal contracts in the first year of the Trump administration's second term, a 70% increase year-on-year. Its Maven Smart System command and control system is about to receive "official project records" status - a valuable recognition in the defense contract field, indicating long-term stable funding security. However, even within the Pentagon, Palantir's absolute dominance is quietly loosening. According to executives from AI startups, the Pentagon is expanding AI deployment from headquarters to the front lines, and lightweight models designed for soldier smartphones or drones are often incompatible with the Palantir system. Palantir has begun to introduce a new version of Maven adapted for drones, but how far this catch-up can go remains unknown. Ben Van Roo, co-founder of AI defense startup Legion Intelligence, pointed out that Maven has been successful, but it covers "only a subset of thousands of workflows within the Department of Defense". More extensive battlefield scenarios such as intelligence collection, logistics support, etc., will give rise to a large demand for AI outside of the Palantir system. He said: That is the main battlefield for the next decade. This article is a reprint from Wall Street News, edited by GMTEight: Chen YuFeng.