Palantir Technologies (PLTR) has been one of the hottest names in the AI rally of 2025, riding a wave of enthusiasm that pushed the stock up more than 140 percent year-to-date and briefly lifting it to an all-time high near $190 just last week. The company, long known for its government contracts and classified data analytics, had reinvented itself in the market’s imagination as a pure-play AI champion. Yet the past several days have delivered a sharp reversal, with shares tumbling close to 10 percent on August 19 and extending a five-session losing streak. The catalyst was a blunt note from Citron Research, one of Wall Street’s most notorious short-selling outfits, which claimed Palantir’s valuation is “detached from fundamentals” and unsustainable in light of peers like OpenAI.
Citron’s Andrew Left drew a sharp contrast between Palantir and OpenAI, pointing to valuation multiples as the core of his bearish case. OpenAI, now widely considered the benchmark for commercial AI growth, trades at roughly 17 times forward revenue. Palantir, by contrast, commands an extraordinary 90 times multiple, according to Citron’s calculations. If one were to apply the OpenAI ratio to Palantir’s projected 2026 revenue, the implied share price would collapse to around $40. Even under generous assumptions, Citron argued, the market has inflated Palantir far beyond what its revenue base and commercial traction can currently justify. This argument resonated with investors, many of whom have been quietly wary of how fast AI-related stocks had run.
The impact was immediate. The selloff not only erased billions in market value but also underscored how sensitive Palantir has become to shifts in sentiment. While bulls argue the company has a unique positioning, with both deep U.S. government contracts and growing commercial adoption of its Foundry and AI Platform, skeptics view it as a business still struggling to scale beyond its defense-heavy roots. Traditional analysts also lean cautious: of eight recently surveyed, just two call Palantir a “Buy,” with most preferring “Hold” ratings. Their average target sits in the mid-$150s, reflecting the belief that the stock is neither a screaming bargain nor a stable safe haven at these levels.
Yet there are contrasting voices. TipRanks’ AI-driven forecasting model has awarded Palantir a robust “Outperform” score in the mid-70s, projecting that shares could recover and potentially reach $196. Bulls argue that as AI adoption accelerates, Palantir’s ability to operationalize large-scale data for defense and enterprise customers sets it apart from startups that lack proven deployment pipelines. They also emphasize Palantir’s consistent profitability, a rarity among AI-themed software stocks, as evidence that the company has staying power even if multiples compress.
The real test for Palantir now lies ahead. On one side stands the bearish thesis—that the company has been swept up in an AI bubble, trading on hype and scarce comparables, destined for a painful re-rating. On the other side is the bullish view—that Palantir has the right technology, contracts, and financial discipline to grow into its premium valuation as AI becomes a pervasive enterprise tool. Much will depend on whether Palantir can translate its government-focused strength into scalable commercial deals that drive revenue growth without sacrificing margins.
For investors, the debate mirrors the broader question gripping the market: are AI valuations an echo of the dot-com bubble, or the beginning of a new long-term productivity cycle? Palantir is now the latest test case, caught between hype and execution. If it delivers accelerating adoption in the private sector, the bulls may yet win. If not, Citron’s $40 target could look less like hyperbole and more like a sobering warning.