Close Menu
GeekBlog

    Subscribe to Updates

    Get the latest creative news from FooBar about art, design and business.

    What's Hot

    Grok Says the Scariest Thing in the Universe Is False Vacuum Decay. Here’s What Physicists Actually Think.

    November 16, 2026

    Xbox Says Project Helix Is a Console That Plays Your PC Games, Not a PC That Plays Xbox Games

    November 16, 2026

    The FAA Grounded Every US Flight for 4 Hours. Duffy Says It Will Happen Again Unless Congress Pays.

    November 16, 2026
    Facebook
    GeekBlog
    • Home
    • Mobile
    • Tech News
    • Blog
    • Gaming
    • Smartwatch
    • How-To Guides
    • AI & Software
    Facebook
    GeekBlog
    Home»AI & Software»The NSA Just Named Six Chinese AI Companies. It Says They Have Been Draining Claude, GPT and Gemini Since 2024.
    AI & Software

    The NSA Just Named Six Chinese AI Companies. It Says They Have Been Draining Claude, GPT and Gemini Since 2024.

    Marcus BennettBy Marcus BennettSeptember 10, 20267 Mins Read
    Share Facebook Twitter Pinterest LinkedIn Tumblr Email Copy Link
    Illuminated data center server racks representing AI model infrastructure
    Photo via Pexels
    Share
    Facebook Twitter LinkedIn Pinterest Email Copy Link

    On September 8, the National Security Agency, the Cybersecurity and Infrastructure Security Agency, and the FBI published a joint cybersecurity advisory that reads less like a typical breach notice and more like an indictment of an entire national strategy. Advisory AA26-251A names six Chinese AI companies, DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun and Z.AI, and accuses them of running industrial-scale campaigns to siphon the outputs of America’s most advanced AI models and use them to train their own.

    The three agencies say the activity has been running since at least late 2024, involves millions of individual requests, and has pulled billions of tokens out of frontier systems built by Anthropic, OpenAI, Google and xAI. Just as notable as the accusation itself is the advice the agencies are giving American AI companies in response: quietly feed suspected offenders a worse, cheaper model, and do not tell them it happened.

    Quick facts

    • Advisory AA26-251A was published jointly by the NSA, CISA and the FBI on September 8, 2026
    • It names DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun and Z.AI
    • The agencies say the campaign has run since at least late 2024, likely with Chinese government awareness
    • Billions of tokens were extracted across millions of requests from Claude, GPT, Gemini and Grok models
    • Z.AI is singled out for distilling billions of tokens of GPT-5.5 and Claude Opus 4.8 data by mid-2026
    • Recommended response includes silently serving “downgraded” models to suspected accounts

    What distillation is, and where it turns into theft

    Knowledge distillation is not, on its own, anything unusual or illegal. It is a standard machine learning technique where a smaller “student” model learns by studying the outputs of a larger, more capable “teacher” model, a shortcut that can save enormous amounts of training cost and time. Every major AI lab uses some version of it internally, and plenty of legitimate research relies on studying published outputs from other systems.

    What the advisory describes is different in kind, not just degree. Instead of distilling from data they own or are licensed to use, the agencies say the six named companies ran millions of automated queries against competitors’ commercial APIs, systematically harvesting responses at a scale and pattern far outside normal usage, then used that harvested output to train and improve their own models. That shifts the activity from a legitimate research shortcut into what the advisory treats as unauthorized extraction of a competitor’s product, executed through the front door of a public API rather than through any kind of technical break-in.

    Recommended for you:

    OpenAI Says Its Next Model Solved a $1 Million Math Problem. A Rival Lab’s Researchers Say They Got There First.
    AI & Software·Sep 10, 2026

    OpenAI Says Its Next Model Solved a $1 Million Math Problem. A Rival Lab’s Researchers Say They Got There First.

    How the agencies characterized it

    The advisory describes the campaign as “aggressive, malicious, and targeted,” and assesses that its scale and coordination make it likely the Chinese government is aware of the activity, treating rapid distillation of foreign frontier models as a core part of these companies’ development strategy rather than an occasional shortcut.

    Who the advisory says targeted what

    The six companies were not accused uniformly. The advisory draws a rough hierarchy, with DeepSeek and Moonshot AI named as the most aggressive offenders, distilling from multiple Claude, GPT, Gemini and Grok models at once. The remaining four firms are described as more selective, each concentrating on a narrower set of targets.

    CompanyModels reportedly targetedAdvisory’s characterization
    DeepSeekClaude, GPT, Gemini and Grok modelsNamed a top offender
    Moonshot AIClaude, GPT, Gemini and Grok modelsNamed a top offender
    MiniMaxClaude, Gemini and GPT modelsBroad but narrower targeting
    AlibabaClaude and GPT modelsTargeted improvement of own products
    StepFunClaude and GPT modelsTargeted improvement of own products
    Z.AIGPT-5.5 and Claude Opus 4.8 specificallyBillions of tokens distilled by mid-2026 to build reasoning capability

    The Z.AI detail is the most concrete number in the whole advisory, and it is worth sitting with. Agencies say that by mid-2026 the company had distilled billions of tokens worth of output specifically from GPT-5.5 and the kind of frontier reasoning models labs like OpenAI have been racing to push forward this year, along with Claude Opus 4.8, in order to build up its own model’s chain-of-thought reasoning ability. In plain terms, the advisory is saying Z.AI used the step-by-step thinking of two of the world’s most expensive AI systems as a training set, at a fraction of the cost it would take to develop that reasoning ability independently.

    How a distillation campaign works Three steps, all through a normal-looking API 1. QUERY AT SCALE Millions of automated requests sent to a frontier model’s API, disguised as normal traffic 2. HARVEST OUTPUTS Responses and reasoning traces are logged and cleaned into a training dataset 3. TRAIN THE STUDENT A cheaper model learns to imitate the teacher’s behavior, at a fraction of the R&D cost The advisory says the same technique used legitimately inside a single lab becomes theft when it targets a rival company’s API without authorization, at industrial scale, over roughly two years.
    Illuminated data center server racks running active AI infrastructure equipment

    Every one of the millions of queries described in the advisory passed through ordinary-looking API infrastructure like this, not a break-in. Photo via Pexels.

    The part that will bother privacy advocates: quiet downgrades

    The advisory does not stop at description. It tells American AI companies how to respond, and the recommended playbook is unusually blunt for a government document. Providers are told to improve behavioral and infrastructure-level detection of distillation-style query patterns, and when an account is suspected, to respond not by banning it outright but by quietly routing it to a less capable, “downgraded” version of the model instead of the real one.

    The advisory goes further and explicitly instructs providers to avoid telling suspected accounts that this switch has happened. The stated logic is straightforward: notifying someone that they have been detected only teaches them how to adjust their query patterns and evade detection next time. It is also the kind of instruction that sits uneasily next to normal expectations of transparency, since it means any customer, including ordinary businesses and developers with no connection to a distillation campaign, has no way to know for certain whether an unusually weak or inconsistent answer from a commercial AI API reflects a real model limitation or a silent, unannounced downgrade triggered by an automated fraud system that got a false positive.

    Why now, and why it matters beyond one advisory

    Recommended for you:

    Is PayPal Down? How to Check PayPal and Venmo Status and Fix It Fast
    AI & Software·Sep 9, 2026

    Is PayPal Down? How to Check PayPal and Venmo Status and Fix It Fast

    The timing lines up with a broader pattern this year of Chinese AI firms closing the gap with US labs on both price and capability. DeepSeek in particular slashed its own prices to near nothing while investors valued the company at $74 billion anyway, a business model that becomes far easier to sustain if a meaningful share of your model’s underlying reasoning ability was trained on another company’s far more expensive research rather than built from scratch. The same frontier labs the advisory says were targeted have also been treating their own most capable systems as sensitive assets in other contexts, restricting access to advanced capabilities the way OpenAI limited its Astra model’s most powerful cybersecurity tools to a small vetted group rather than the general public. Extraction at the scale described in this advisory threatens the entire economic logic behind spending hundreds of millions of dollars training a frontier model in the first place, since that advantage erodes quickly if a rival can approximate the result for a fraction of the cost by querying the finished product instead of building their own from the ground up.

    What happens next

    None of the six named companies had issued a formal public response to the specific advisory at the time of writing. The agencies have not announced sanctions or legal action tied to the report, framing it instead as guidance for industry rather than the opening move in an enforcement case. For the American AI companies on the receiving end, the practical choice now is between two uncomfortable options: tighten detection and start quietly degrading service for suspected accounts, risking legitimate customers getting caught in the net, or leave their APIs open as before and accept that a growing share of their most expensive research may keep ending up, one query at a time, inside a rival’s next model release.

    Sources and further reading

    • CISA: Advisory on China based AI firms running distillation campaigns
    • CISA: CISA, NSA and FBI warn of industrial scale AI model distillation
    • Engadget: US accuses Chinese AI firms of copying American models

    About this article: GeekBlog covers U.S. technology news, AI, phones, smartwatches and gaming. Every story is written and checked under our Editorial Policy. Spotted a mistake or have a story tip? Contact our editors.

    AI Anthropic China OpenAI Security
    Share. Facebook Twitter Pinterest LinkedIn Tumblr Telegram Email Copy Link
    Previous ArticleOpenAI Says Its Next Model Solved a $1 Million Math Problem. A Rival Lab’s Researchers Say They Got There First.
    Next Article Samsung Told Apple to Stop Reheating Its Leftovers, and It Has a Seven Year Head Start
    Marcus Bennett

      Marcus Bennett is GeekBlog's Android expert, covering everything from Google's Pixel line and Samsung Galaxy flagships to OnePlus, Nothing, Xiaomi and the broader Android ecosystem. He follows each Android OS release, One UI and Pixel Feature Drop, custom ROMs and the foldable wave, translating spec sheets and beta builds into hands-on guidance for readers choosing their next Android phone, tablet or wearable.

      Related Posts

      9 Mins Read

      Grok Says the Scariest Thing in the Universe Is False Vacuum Decay. Here’s What Physicists Actually Think.

      9 Mins Read

      The FAA Grounded Every US Flight for 4 Hours. Duffy Says It Will Happen Again Unless Congress Pays.

      8 Mins Read

      Apple’s Health AI Study Wants 200,000 iPhone Users. Here’s Who Qualifies and What You Share.

      7 Mins Read

      A Stolen Fireworks AI Key Let Hackers Backdoor Claude Code in Cursor. Here’s How to Check Your Machine.

      6 Mins Read

      Apple May Kill the iPhone’s Action Button in 2027. Camera Control Would Take Over Its Job.

      6 Mins Read

      NASA Will Pay Students a Share of $1 Million to Design Its Moon Base. Here’s How to Enter.

      Top Posts

      Every iPhone Camera Ranked in 2026 (Best to Worst)

      July 6, 2026361 Views

      Best Browser for Android in 2026: Top Picks Ranked

      July 7, 2026268 Views

      Apple Watch Black Friday 2026: Early Deals, Prices to Beat and When to Buy

      November 8, 2026177 Views
      Stay In Touch
      • Facebook

      Subscribe to Updates

      Get the latest tech news from FooBar about tech, design and biz.

      Most Popular

      MapQuest Now Shows the Plate Readers on Your Route. It Won’t Say Where the Map Comes From.

      October 19, 2026660 Views

      Apple Told Its Stores Not to Expect iPhone Duo Walk-Ins. Carriers Can Beat January by Weeks.

      October 20, 2026312 Views

      Apple Watch Black Friday 2026: Early Deals, Prices to Beat and When to Buy

      November 8, 2026216 Views
      Our Picks

      Grok Says the Scariest Thing in the Universe Is False Vacuum Decay. Here’s What Physicists Actually Think.

      November 16, 2026

      Xbox Says Project Helix Is a Console That Plays Your PC Games, Not a PC That Plays Xbox Games

      November 16, 2026

      The FAA Grounded Every US Flight for 4 Hours. Duffy Says It Will Happen Again Unless Congress Pays.

      November 16, 2026

      Subscribe to Updates

      Get the latest creative news from FooBar about art, design and business.

      HEICJPG.online - Convert HEIC to JPG online
      Facebook
      • About GeekBlog: Our Mission, Team and Editorial Standards
      • Contact us
      • Privacy Policy
      • Disclaimer
      • Terms and Conditions
      • Editorial Policy
      • Cookie Policy
      • Your Privacy Choices
      © 2026 GeekBlog

      Type above and press Enter to search. Press Esc to cancel.