⚡️⚡️🤖 BREAKING: All Western AI Stolen by China

🔥🪖 🇨🇳⚔️🇺🇸 Fox, WSJ, Reuters, Bloomberg simultaneously published on a Sunday evening: "#China stole #Claude!"

OpenAI immediately chimed in: "Us too! Save us!"

"We identified large-scale campaigns conducted by three AI labs — DeepSeek, Moonshot, and MiniMax — aimed at illegally extracting Claude's capabilities to improve their own models," Anthropic obediently admitted in a statement published under tomorrow's date.

According to Anthropic's report dated February 24 — which Fox News Digital received (or wrote) first on February 23 — DeepSeek, Moonshot AI, and MiniMax allegedly used approximately 24,000 fraudulent accounts to generate over 16 million transactions through Claude as part of a "coordinated large-scale distillation campaign aimed at extracting valuable data."

❗️"Anthropic has consistently supported export controls to help maintain America's AI leadership," CEO Amodei pleads in his "report."

So let me get this straight — they're saying: "We provided Claude's services for training Chinese models, but it's not our fault we got fooled again"?

Let's recall: in November 2025, a week before the Claude Opus 4.5 release, Anthropic similarly claimed Claude was compromised via social engineering by a group called "GTG-1002" — which, as it turned out, doesn't exist.


🔴 Timeline

• Feb 5 — Anthropic and OpenAI simultaneously, to the minute, on cue from Russian-born Israeli-Canadian citizen Ilya Sutskever, released GPT-5.3-Codex and Claude Opus 4.6

• Feb 3–9 — Safety researchers flee Anthropic en masse, led by Head of Safety Mrinank Sharma

• Feb 12 — Google claims hackers are using Gemini AI at every stage of cyberattacks

• Feb 12 — OpenAI sends memo to House Select Committee on China: "DeepSeek is distilling our models"

• Feb 13 — OpenAI discontinues ChatGPT support for GPT-4o, GPT-4.1, GPT-4.1 mini, and o4-mini, and "decommissions" GPT-5 (Instant and Thinking)

• Feb 14 — Hegseth and OpenAI announce that the Chinese-stolen GPT (decommissioned the day before) will be urgently integrated into GenAI.mil alongside previously compromised Gemini 3 Pro already integrated into the Pentagon

• Feb 17 — Palantir moves HQ from Denver to Miami "tax haven" — same day releases report showing CEO Karp racked up $17.2M in Asia flights in 2025

• Feb 19 — Hegseth's Under Secretary for R&E Emil Michael: "Cross the Rubicon"

• Feb 20 — Anthropic rolls out Claude Code Security — "we know how to find vulnerabilities"

• Feb 23 (Sunday) — Fox, WSJ, Reuters, Bloomberg in sync: top intelligence analyst kidnapped by China along with everyone else

• Feb 23/24 — Anthropic: "Yes sir, everything was stolen before we got here!"

• Feb 24 — Hegseth–Amodei meeting

A full report on the leaks and more will likely be tomorrow. After the meeting, for which everyone prepared so meticulously that they forgot to check the dates of their own press releases.


➡️ Two scenarios — both worse

• Scenario A: "Give us money"
All companies simultaneously craft a narrative: "China is stealing American AI." Goal: increased funding, tighter export controls (plays into Trump's hand), new government contracts on home turf.

• Scenario B: "Nationalize us"
All companies simultaneously demonstrate they can't protect their models from China. Why? To sell more to the potential adversary?

Subscribe | Send content | Boost