On the day Palantir Technologies set a fresh record on the stock market, its chief executive was talking about plumbers, welders, and people whose brains work differently from the norm. Alex Karp did not open with a boast about the company’s run. He opened with a theory about who wins as artificial intelligence spreads through the economy.
The shares gave him the platform. Palantir closed Friday at $209.05, up 5.17%, topping the previous record close of $207.18 set in November 2025. The move followed Barclays’ decision to initiate coverage with an Overweight rating and a $265 price target, a day after Goldman Sachs upgraded the stock to Buy with a $230 target and Raymond James moved to Outperform. Wall Street, after months of caution, had turned loud about the company again.
Speaking at a Fortune event the previous day, Karp argued that only two kinds of people are safe in the AI era: those with vocational training in a skilled trade, and the neurodivergent. He told a younger audience to skip elite college degrees, and he aimed a jab at rival Anthropic over its hiring and its safety messaging. Karp, who is dyslexic and has long cast his own wiring as an asset, was restating a familiar theme: the technology Palantir sells will reward the people who are least interchangeable with a machine.
The stock’s climb carries a harder question than Karp’s career advice. Palantir now trades near 170 times earnings, well above the roughly $199 consensus price target, and analysts keep raising numbers faster than the company books orders. Barclays’ $265 target rests on a bet that sovereign AI demand, governments and defense agencies buying its software to run their own models, keeps compounding, according to analysts.
That bet has been paying off. Palantir has spent the year outperforming most software peers, powered by the AIP platform that wraps its data tools around large language models, and by an expanding commercial book that reaches beyond its long-time government base. The company reports third-quarter results in November, and the record share price raises the bar for what the numbers must show.
Palantir’s growth now comes from two directions at once. The government business, which built the company, has expanded beyond intelligence into defense, health, and energy agencies, and the rise of sovereign AI has turned public budgets into a fresh source of demand. The commercial business, smaller but faster-growing, sells to manufacturers, airlines, and hospitals through those bootcamps, which executives credit with shortening sales cycles that once dragged on for a year or more.
The upgrades also arrived during a stretch when software shares were recovering as a group. Analysts who raised their targets pointed to the same underlying argument: Palantir sits in the middle of the trade that has defined the past two years, in which investors pay up for companies positioned to profit from AI spending rather than simply consume it.
Karp has never been shy about the valuation. He has argued for years that the company is priced as a conviction, not a spreadsheet, and that the market is still catching up to how deeply AI will run through defense and intelligence budgets. His remarks about trades and neurodiversity fit the same persona: a contrarian who frames every debate, hiring, safety, pricing, as a test of who adapts and who does not.
The Fortune comments also landed in a week when AI companies were trading accusations about runaway systems. Karp pointed at Anthropic’s hiring and safety posture, a jab that had little to do with Palantir’s quarterly numbers and everything to do with positioning. Palantir sells to institutions that want control over their own AI, and Karp has made that contrast with model-makers a recurring theme.
The rally has an echo of earlier spikes. Palantir has surged before on analyst enthusiasm, only to give gains back when orders or guidance disappointed. The difference now is the breadth of the endorsement, three upgrades in as many days, and the size of the gap between where the stock trades and where most analysts still think it belongs.
The company was founded in 2003 by a group that included Karp and the investor Peter Thiel, and it takes its name from the seeing-stones in J.R.R. Tolkien’s fiction. For most of two decades it sold to intelligence agencies and militaries, software built to find patterns in sprawling, unruly data. The commercial push that now drives the growth story began in earnest around 2023, when Palantir started running its bootcamps, hands-on sessions that handed prospective customers working prototypes in days rather than quarters.
What Palantir must now deliver is the growth that the new targets assume. Karp’s message to young people was simple: learn a trade, or lean into a different kind of mind. The message to his own investors is less settled. The stock has priced in a future in which sovereign demand compounds; the November report will show whether the orders are arriving fast enough to justify the record.
