ainewsblitz.com

Breaking

Chinese Token Resellers Offer Unauthorized Claude Access While Labs Extract Model Reasoning

  • Foundation Models
  • Security

A thriving gray market of Chinese resellers is circumventing Anthropic's access restrictions by offering Claude API tokens at discounts of 70–90% below official prices, enabling undisclosed model-distillation campaigns that have extracted nearly 200 million exchanges of user interactions. The operations, documented by The Information and detailed in Anthropic's September 2026 threat report, illustrate a persistent cat-and-mouse dynamic: as Anthropic tightens controls, resellers develop new workarounds ranging from stolen credentials to shell companies to overseas subsidiaries.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.

$20
Read this article
$29/month
Unlimited — all 26,886 articles, the full archive, and comprehension quizzes
Save 72%
$98/year
≈ $8.17/month
Unlimited, billed once a year