Palantir

Palantir Technologies sells software that fuses scattered data into one operating picture — for governments first, for companies later. Peter Thiel and Alex Karp, among others, founded it in 2003 to build case-management tools for post-9/11 intelligence work. It stayed private for 17 years, raising money at a $15B valuation in 2015, per a Wall Street Journal report at the time, then went public through a direct listing on the New York Stock Exchange on September 30, 2020; CNBC put its first-day close at $9.50 a share. It moved its listing to Nasdaq in November 2024, the same year it joined the S&P 500. By July 2025, CNBC ranked it the 20th most valuable company in the US.

That path — spy-tech contractor to Nasdaq mainstay — is why Palantir carries more argument per dollar of revenue than almost any other large-cap name. It has no clean peer. It sells to the same government agencies that critics accuse it of arming and to pharmaceutical companies from the same platform. Its CEO writes books and picks public fights instead of avoiding them. Sell-side analysts, national-security reporters, and retail forums all cover the same four letters with almost nothing else in common. The result is a stock discussed in several unrelated vocabularies at once, each certain the others are missing the point.

This report doesn’t add a fourth vocabulary or a verdict. It measures the three that already exist, using twelve years of coverage TEXXR holds in full.

Three periods show up cleanly in the record.

Govtech and the surveillance question (2015–2020). Coverage of this period ran two threads in parallel: financing news, and accountability reporting on what the software did. Wired described Palantir building “an opaque surveillance network” from police-department data; separate reporting documented that ICE used Palantir’s case-management tools to locate and arrest people. Palantir carried that combination — quiet funding rounds, loud scrutiny — into its 2020 direct listing, which CNBC greeted by calling it “Big Data’s scariest, most secretive unicorn.”

The AIP re-rating (2023–present). Palantir shipped its Artificial Intelligence Platform in 2023. TEXXR’s full edge list does not show coverage turning more financial after it. Financial-predicate edges — revenue, contract value, share price — made up 44% of Palantir’s edges in 2024 (21 of 48) and 31% in 2025 (23 of 75). TEXXR changed how it extracts edges in early 2026, so later shares do not compare. Revenue growth accelerated all the same: 18% year-over-year in Q1 2023, 70% in Q4 2025, then 85% in Q1 2026, with US commercial revenue up 133%, per Palantir’s own quarterly reports as covered by CNBC and Reuters.

That gap is the corpus-only fact this report is built to surface: as revenue growth rose from 18% to 70% by late 2025, the money share of Palantir’s edges fell 13 points and the controversy share quadrupled. The record did not settle into the vocabulary of a stock; it went back to what the software does, and for whom.

Same contract, three vocabularies. Palantir’s UK National Health Service relationship shows the mechanism up close, inside a single, mostly unchanged deal. In 2022, Bloomberg described “a secret plan to deepen its ties” to the NHS. By late 2023 that same relationship was a contract worth up to £480M, reported plainly. By mid-2026, with the underlying arrangement barely altered, Reuters reported a government review of the contract and the Financial Times ran a piece on “Palantir’s fight to save” it — while London’s mayor separately blocked a £50M Palantir deal with the Metropolitan Police over what his office called a “clear and serious breach” of procurement rules, per the Guardian. One relationship, three vocabularies, four years.

From 2020 through 2025, under one extraction method, Palantir’s edge record was led by financial edges (42% of 221), ahead of partnership (17%) and controversy (14%). TEXXR changed how it extracts edges in early 2026, so later shares are not comparable with these. Within the earlier method, controversy was the thread that grew: 4% of edges in 2024 (2 of 48), 16% in 2025 (12 of 75).

Two May 2026 edges describe “frontier AI labs” — the model builders, not the systems integrators Palantir used to be grouped with — pressing on its position, a month before CEO Alex Karp told CNBC that Palantir’s enterprise customers were “unhappy” with how those labs operate, and weeks later called US military reliance on AI labs “effing insane” in Forbes. Palantir sits downstream of the AI capex buildout rather than inside it — it doesn’t build chips or foundation models — but the graph narrates it alongside the companies that do, including a July 2026 deal to deploy Nvidia’s Nemotron models for US agencies, a thread that ties Palantir’s edge record directly to Nvidia’s own coverage arc.

The controversy edges haven’t receded under the financial re-rating. They’ve moved location. Internal Slack logs showing Palantir staff debating the company’s ICE and Department of Defense work became coverage in April 2026, per Ars Technica; by July, the Financial Times quoted insiders and investors worrying that the company’s alignment with the Trump administration could cost it corporate clients and engineering talent.

Positioning: how much of PLTR is sold short, on the record brokers file twice a month. It lags by design — read it as where the crowd stood, not where it stands.

Signals in the record worth tracking, not trade ideas:

• Whether frontier labs keep appearing as rivals. Two May 2026 edges cast them that way. TEXXR’s “competitor” label exists only since its 2026 extraction change, so count such edges rather than the label’s share. Frontier labs appearing as rivals rather than customers or partners is a measurable shift in the graph, independent of any single quarter’s revenue.
• The outcome of the NHS contract review and the Met Police block. The record shows no outcome for either as of 27 October. A contract termination or reversal would register as a hard status change in the edge record, not a shift in tone.
• Whether controversy keeps gaining on financial edges, measured within TEXXR’s 2026 extraction rather than against 2025 — the two have traded places before in this same graph, and the re-rating is not the only pattern TEXXR has recorded here.
• Karp’s public commentary volume. His remarks generate a disproportionate share of Palantir’s controversy and personnel edges; a change in his public posture tends to show up in the predicate mix before it shows up in a press release.

Palantir lives or dies by the same constraint as these names — small equal-weighted groups, one constraint each. Not investment products: see all categories.