The Short-Lived White House Press Secretary and the Mining AI That Never Was
By Tess Owen and Tim Marchman | October 16, 2026
The short-lived White House press secretary was listed as an adviser to a Kristi Noem-linked company that aimed to use AI for mining. Both the company and she say she never worked there.
In an era when artificial intelligence is being applied to everything from drug discovery to defense logistics, a curious case has emerged at the intersection of AI, critical minerals, and political personnel. It involves Katie Zacharia — a figure whose tenure as White House press secretary was notably brief — and a startup claiming to harness AI for rare-earth mining, allegedly connected to a political network tied to Kristi Noem.
The problem? According to both Zacharia and the company itself, she never worked there.
A Startup That Says It Uses AI to Mine Rare Earths
The company in question positioned itself in one of the most strategically sensitive sectors of the 2026 economy: the extraction and processing of rare-earth elements. These minerals — essential to semiconductors, electric vehicles, wind turbines, and advanced weapons systems — have become a focal point of US industrial policy as Washington seeks to reduce dependence on Chinese supply chains.
The startup's pitch was straightforward, if ambitious: apply machine learning and AI-driven exploration tools to identify and extract rare-earth deposits more efficiently than traditional geological surveying. It's a thesis that has attracted billions in venture funding across the sector in recent years, with AI-assisted mineral exploration now a recognized — if still nascent — field.
The Kristi Noem Connection
The company was reportedly linked to a network associated with Kristi Noem, whose political profile has remained prominent in 2026. The nature of that connection — whether through investment, advisory roles, or board membership — has not been fully detailed, but it was enough to draw scrutiny when Zacharia's name surfaced as an adviser.
Zacharia's time in the White House was short-lived, and her post-administration activities have attracted outsize attention relative to her tenure. The listing of her name as an adviser to a politically connected AI mining startup thus raised immediate questions about revolving-door dynamics and the increasingly blurred lines between political personnel and private-sector AI ventures.
Both Sides Deny the Connection
In a twist that makes the story stranger still, both Zacharia and the company have stated that she never actually worked there. The discrepancy between the public listing and the denials introduces uncomfortable questions: Was the advisory credit aspirational? Erroneous? A marketing tactic? Or a data-entry artifact in an era when AI-generated company profiles and automated networking databases routinely assemble connections that don't exist in the real world?
The latter possibility is particularly resonant in 2026. As AI tools scrape, infer, and auto-generate professional biographies, false affiliations have become a growing problem — one that can inflate a startup's credibility, complicate a public figure's reputation, and leave journalists and regulators chasing connections that were never real to begin with.
Why It Matters
The Zacharia case is a small but illustrative example of a larger 2026 trend: the convergence of AI hype, critical-minerals geopolitics, and political networking, all operating in a media environment where automated systems can manufacture apparent associations faster than anyone can verify them.
Rare-earth mining is serious business, backed by genuine national-security imperatives. AI is a genuinely transformative — if frequently oversold — technology. And political connections remain as valuable as ever. When all three intersect, the result is a story where the facts are murky, the denials are mutual, and the strangeness is the point.
For now, the question of how Katie Zacharia's name ended up attached to an AI rare-earths startup — and why both she and the company insist it was never true — remains unanswered. In 2026, that may be the most telling detail of all.
via Wired AI