Treasury threatens sanctions after White House claims Moonshot distilled Anthropic’s Fable 66%

By Rebecca Bellan66%

7/22/2026, 8:49:03 PM

BS Summary: This article contains 21 faulty reasoning types, including Appeal to Authority, Confirmation Bias, and Recency Bias, with Pessimism Bias as the most egregious example at 16.3% saturation with 60 hits. Analysis detected 725 faulty-reasoning hits from 368 analyzed words, generating a BS Score of 59.8% and a BS Rank of 66% (7,355 of 21,176 articles). This article is worse (more manipulative) than 65.30% of the article peer group.

U.S. 
Treasury secretary Scott Bessent doubled down on his warnings to Chinese AI companies on Wednesday, saying that sanctions remain on the table after a White House official accused Moonshot of improperly distilling Anthropic’s Fable model. 
Model distillation is a common AI training technique in which a smaller model learns from the outputs of a larger one. 
While this process can infringe on intellectual property rights, it’s also widely used as a legitimate optimization method. 
“Open source is not open season on American IP,” Bessent posted on X. 
“When [Chinese] firms conduct covert, industrial-scale distillation attacks that cross the line into IP theft, sanctions and Entity List designations will be on the table.” 
Earlier this week, Bessent stated that the U.S. government would examine open source models from China for signs of intellectual property theft and impose sanctions if found. 
Bessent's latest remarks come hours after the White House's science and technology policy chief Michael Kratsios accused the China-based Moonshot of conducting large-scale distillation against U.S. models. 
He alleged that Moonshot had acquired Nvidia’s “GB300-equipped servers and has accessed GB300s in Thailand, likely to train its AI models,” raising questions about whether the firm violated U.S. export-control rules. 
The GB300 servers are part of Nvidia’s Blackwell generation, which are banned from being sold to Chinese companies. 
Some experts dispute the idea that Kimi K3 could have been developed primarily through distillation from Fable, which has only been publicly available since July 1. 
Moonshot released K3 last week as an open-weight model, and its advanced capabilities have called into question the underlying business models of leading U.S. 
AI labs, casting doubt on whether they can continue to justify the enormous capital requirements underpinning the frontier AI race. 
The episode has also intensified a broader debate in Washington over the influx of Chinese open models. 
Some, including former White House AI adviser and current OpenAI Head of Strategic Futures, Dean Ball, have argued that the U.S. should restrict or effectively ban the use of Chinese open-weight models to preserve America’s technological advantage and mitigate potential national security risks. 
TechCrunch has reached out to Moonshot and the Treasury for comment. 
Article reasoning-pattern comparisonThis article: 13.9%Rebecca Bellan: 4.5%TechCrunch: 3.0%Confirmation Bias13.9%This article: 8.4%Rebecca Bellan: 1.7%TechCrunch: 1.4%Anchoring Bias8.4%This article: 7.3%Rebecca Bellan: 4.3%TechCrunch: 3.5%Availability Heuristic7.3%This article: 0.0%Rebecca Bellan: 1.6%TechCrunch: 1.1%Representativeness Heuristic0.0%This article: 6.5%Rebecca Bellan: 0.7%TechCrunch: 0.6%Hindsight Bias6.5%This article: 5.4%Rebecca Bellan: 2.2%TechCrunch: 2.5%Overconfidence Bias5.4%This article: 7.9%Rebecca Bellan: 8.3%TechCrunch: 4.8%Framing Effect7.9%This article: 11.7%Rebecca Bellan: 1.6%TechCrunch: 0.6%Loss Aversion11.7%This article: 0.0%Rebecca Bellan: 1.6%TechCrunch: 0.6%Status Quo Bias0.0%This article: 0.0%Rebecca Bellan: 1.0%TechCrunch: 0.2%Sunk Cost Effect0.0%This article: 0.0%Rebecca Bellan: 4.4%TechCrunch: 5.0%Optimism Bias0.0%This article: 16.3%Rebecca Bellan: 3.7%TechCrunch: 1.2%Pessimism Bias16.3%This article: 9.0%Rebecca Bellan: 8.1%TechCrunch: 5.0%Negativity Bias9.0%This article: 0.0%Rebecca Bellan: 4.4%TechCrunch: 2.1%Self-Serving Bias0.0%This article: 8.4%Rebecca Bellan: 0.3%TechCrunch: 0.5%Fundamental Attribution Error8.4%This article: 0.0%Rebecca Bellan: 0.0%TechCrunch: 0.1%Actor-Observer Bias0.0%This article: 0.0%Rebecca Bellan: 1.2%TechCrunch: 0.6%In-Group Bias0.0%This article: 11.7%Rebecca Bellan: 2.4%TechCrunch: 0.3%Out-Group Homogeneity Bias11.7%This article: 0.0%Rebecca Bellan: 1.5%TechCrunch: 3.3%Halo Effect0.0%This article: 0.0%Rebecca Bellan: 0.0%TechCrunch: 0.1%Horn Effect0.0%This article: 0.0%Rebecca Bellan: 0.0%TechCrunch: 0.0%Dunning-Kruger Effect0.0%This article: 12.0%Rebecca Bellan: 2.2%TechCrunch: 2.3%Recency Bias12.0%This article: 0.0%Rebecca Bellan: 0.2%TechCrunch: 0.3%Primacy Effect0.0%This article: 0.0%Rebecca Bellan: 0.0%TechCrunch: 0.1%Blind-Spot Bias0.0%This article: 0.0%Rebecca Bellan: 0.0%TechCrunch: 0.3%Ad Hominem0.0%This article: 0.0%Rebecca Bellan: 0.3%TechCrunch: 0.6%Straw Man0.0%This article: 15.5%Rebecca Bellan: 7.6%TechCrunch: 4.3%Appeal to Authority15.5%This article: 11.7%Rebecca Bellan: 3.2%TechCrunch: 1.7%False Dilemma11.7%This article: 0.0%Rebecca Bellan: 2.1%TechCrunch: 0.6%Slippery Slope0.0%This article: 0.0%Rebecca Bellan: 0.0%TechCrunch: 0.2%Circular Reasoning0.0%This article: 5.4%Rebecca Bellan: 4.5%TechCrunch: 6.0%Hasty Generalization5.4%This article: 0.0%Rebecca Bellan: 0.0%TechCrunch: 0.2%Red Herring0.0%This article: 0.0%Rebecca Bellan: 1.6%TechCrunch: 1.2%Bandwagon0.0%This article: 0.0%Rebecca Bellan: 2.6%TechCrunch: 2.2%Appeal to Emotion0.0%This article: 0.0%Rebecca Bellan: 0.8%TechCrunch: 0.6%Begging the Question0.0%This article: 0.0%Rebecca Bellan: 4.6%TechCrunch: 2.9%Post Hoc (False Cause)0.0%This article: 0.0%Rebecca Bellan: 0.0%TechCrunch: 0.1%Tu Quoque0.0%This article: 0.0%Rebecca Bellan: 0.2%TechCrunch: 0.5%Burden of Proof0.0%This article: 0.0%Rebecca Bellan: 0.0%TechCrunch: 0.2%Appeal to Nature0.0%This article: 0.0%Rebecca Bellan: 0.3%TechCrunch: 0.3%Composition/Division0.0%This article: 0.0%Rebecca Bellan: 2.5%TechCrunch: 2.4%Anecdotal0.0%This article: 6.8%Rebecca Bellan: 0.2%TechCrunch: 0.1%No True Scotsman6.8%This article: 8.4%Rebecca Bellan: 2.6%TechCrunch: 2.0%Ambiguity (Equivocation)8.4%This article: 0.0%Rebecca Bellan: 0.0%TechCrunch: 0.0%Gambler’s Fallacy0.0%This article: 0.0%Rebecca Bellan: 0.3%TechCrunch: 0.2%Middle Ground0.0%This article: 0.0%Rebecca Bellan: 0.0%TechCrunch: 0.0%Personal Incredulity0.0%This article: 0.0%Rebecca Bellan: 0.0%TechCrunch: 0.1%Special Pleading0.0%This article: 0.0%Rebecca Bellan: 0.0%TechCrunch: 0.1%Genetic Fallacy0.0%This article: 12.0%Rebecca Bellan: 5.9%TechCrunch: 2.0%Unattributed Quote12.0%This article: 3.5%Rebecca Bellan: 0.8%TechCrunch: 0.7%Quote-first Misdirection3.5%This article: 3.5%Rebecca Bellan: 3.0%TechCrunch: 4.6%Biased Writer Voice3.5%This article: 11.7%Rebecca Bellan: 2.4%TechCrunch: 0.8%Indoctrination11.7%This article: 0.0%Rebecca Bellan: 0.0%TechCrunch: 0.1%Politically Left Leaning Bias0.0%This article: 0.0%Rebecca Bellan: 0.0%TechCrunch: 0.1%Politically Right Leaning Bias0.0%This article: 0.0%Rebecca Bellan: 2.8%TechCrunch: 4.6%Attempt to Sell a Product or S…0.0%

368 words analyzed.

Speakers

4speakers55%attributed speech166writer words
Voice mapSelect a segment to jump to its words
Writer's voice • 11 words • 0.0% coverageTreasury • 1 words • 0.0% coverageScott Bessent • 35 words • 0.0% coverageWriter's voice • 21 words • 0.0% coverageWriter's voice • 18 words • 0.0% coverageScott Bessent • 13 words • 100.0% coverageScott Bessent • 25 words • 0.0% coverageScott Bessent • 27 words • 0.0% coverageMichael Kratsios • 27 words • 0.0% coverageMichael Kratsios • 31 words • 100.0% coverageWriter's voice • 18 words • 0.0% coverageWriter's voice • 26 words • 0.0% coverageWriter's voice • 24 words • 0.0% coverageWriter's voice • 20 words • 0.0% coverageWriter's voice • 17 words • 0.0% coverageDean Ball • 43 words • 100.0% coverageWriter's voice • 11 words • 0.0% coverage
Selected voice

Dean Ball

100%flagged-word coverage
43 attributed words21% of attributed speech70% writer coverage
0%50.0%100.0%Indoctrination+100.0 ptsWriter: 0.0%Dean Ball: 100.0%100.0%

Attribution is sentence-level. Pattern percentages are calculated only from words assigned to that voice.

Loading…
Loading…
Loading…

Analysis

Hover over highlighted words in the article to view the associated bias or fallacy analysis.