China NSCC Tianjin Hack: 10 Petabytes of Missile Data Allegedly Stolen and Beijing Has Said Nothing

GigaNectar Team

IBM Blue Gene P supercomputer server racks at Argonne National Laboratory used for large-scale scientific and defense research computing
UNVERIFIED — UNDER REVIEW

A hacker group operating under the name FlamingChina is claiming to have stolen over 10 petabytes of sensitive data from China’s National Supercomputing Center (NSCC) in Tianjin — a state-run facility that serves more than 6,000 clients across China, including defence agencies, aerospace institutions and advanced research organisations. The alleged dataset is reportedly being offered for sale in cryptocurrency, with a preview access priced at thousands of dollars and full access quoted at hundreds of thousands.

The group posted a sample of the claimed dataset on an anonymous Telegram channel on February 6, 2026, stating it contained “research across various fields including aerospace engineering, military research, bioinformatics, fusion simulation and more.” The post linked the stolen data to top organisations including the Aviation Industry Corporation of China (AVIC), the Commercial Aircraft Corporation of China (COMAC), and the National University of Defense Technology.

Cybersecurity experts who reviewed the samples said the data appeared consistent with what a facility of this scale would hold. China’s Ministry of Science and Technology and the Cyberspace Administration of China had not issued any public statement at the time of reporting. For context on the broader AI and technology security race between major powers, related coverage is available. The alleged breach, if confirmed, would rank among the largest data thefts ever attributed to a single state-run facility.

Inside the Alleged 10-Petabyte Heist

How a compromised VPN reportedly turned into a six-month data drain — step by step

10 PB
Alleged volume stolen
6,000+
NSCC clients served
~6 mo.
Claimed extraction time
2009
NSCC Tianjin founded — first in China

Tap or click each step to read the detail

01
Compromised VPN Entry Point
The attacker, as described to cybersecurity researcher Marc Hofer of the NetAskari blog, claimed to have entered the NSCC’s network through a compromised VPN domain — exploiting a misconfigured or credential-exposed access point rather than a sophisticated zero-day exploit. The entry was described as comparatively straightforward.
02
Lateral Spread via Botnet
Once inside, the attacker reportedly deployed a botnet — a network of automated programs — that moved laterally across the NSCC’s storage clusters, progressively gaining access to more infrastructure components over an extended period without triggering detection systems.
03
Distributed Extraction — Small Flows, Many Channels
Rather than routing large volumes of data through a single pipe — which would raise alarms — the attacker distributed the extraction across multiple servers simultaneously. Small, steady data flows to different destinations lowered the chance of triggering anomaly detection. Dakota Cary of SentinelOne described the approach as architectural rather than technically novel: “You can think of it as having a bunch of different servers that you have access to and you’re pulling data through this hole in the security of the NSCC.”
04
Six Months, Undetected
The full extraction of the alleged 10 petabytes reportedly took approximately six months to complete. The NSCC has not publicly confirmed any breach detection or response activity during this window. Chinese authorities issued no statement about the incident during this period.
05
Dark Web Listing & Sample Release
On February 6, 2026, an account named FlamingChina posted sample data on an anonymous Telegram channel. A separate listing under the alias “airborneshark1” also surfaced on dark web forums — initially offering preview access for approximately $3,000 USD in cryptocurrency, with full dataset access quoted at hundreds of thousands of dollars. The re-listing was noted by researchers as likely intended to drive up bidding.
Tianhe-1 supercomputer at the National Supercomputing Center in Tianjin China
Staff members walk past the Tianhe-1 supercomputer at the National Supercomputing Center in Tianjin, China (November 2010). The Tianjin center was the first national supercomputing hub in China when it opened in 2009.
Photo Source: Wikimedia Commons / Public Domain
🛡️
Classified Documents
Documents labeled 秘密 (secret, in Chinese) classified for up to 10 years. A 2025 report included results of bunker-busting munition tests.
🚀
Missile & Weapons Data
Technical files, animated simulations and renderings of defense equipment including missile warhead testing and bomb systems.
✈️
Aerospace Engineering
Research data linked to AVIC and COMAC, including stealth-related files with binary data referencing the Stealth project.
🎯
Target Simulation Models
Virtual simulations of strikes on aircraft carriers and bunker facilities; physical models simulating damage to a HIMARS launcher’s components.
🧬
Bioinformatics & Fusion
Research across bioinformatics and nuclear fusion simulation from institutions including the National University of Defense Technology.
📡
Radar System Test Data
Radar test results alongside complex data tables related to radar system development — part of the broader defense R&D sample set.
1 Laptop (~1 TB)
1 TB
1 Petabyte
1,000 TB
5 Petabytes
5,000 TB
Alleged breach (10 PB)
10,000 TB

1 petabyte = 1,000 terabytes. A high-spec laptop typically holds around 1 terabyte.

“They’re exactly what I would expect to see from the supercomputing center. You would use supercomputer centers for large computational tasks. The swath of samples that the sellers put out kind of really speaks to the breadth of customers that this supercomputing center had.”
Dakota Cary — Consultant, SentinelOne (China-focused cybersecurity)

Background & Prior Incidents

The Tianjin center was the first national supercomputing hub in China when it opened in 2009. It is one of several such facilities across the country, with others operating in Guangzhou, Shenzhen and Chengdu. According to Marc Hofer, only state-level intelligence services likely have the capacity to process the entire 10-petabyte dataset and extract anything operationally useful from it.

This is not China’s first major data exposure. In 2021, a database reportedly containing personal information of up to 1 billion Chinese citizens was left publicly accessible for over a year before it drew attention in 2022, when an anonymous user offered the data for sale. Dakota Cary noted that China’s cybersecurity posture across both government and private sectors has long been considered a structural weak point — a position that Chinese policymakers themselves have publicly acknowledged.

China’s National Security White Paper (2025) listed building “robust security barriers for the network, data, and AI sectors” as a key national priority, alongside continued efforts to strengthen coordinated cybersecurity mechanisms for critical information infrastructure. In the weeks after the alleged breach became public, the Chinese Academy of Sciences reportedly removed several senior officials involved in high-level military R&D — among them Yang Wei, chief designer of the J-20 stealth fighter. No official connection between these removals and the alleged breach has been stated publicly. Those tracking the broader pace of technology and security developments in 2026 have noted this incident as part of a wider pattern of cyber exposure at state-linked facilities.

The alleged breach of China’s National Supercomputing Center in Tianjin was covered in this piece, including the attack method described by the hacker to researcher Marc Hofer, the data categories present in the sample release, the pricing of the dataset on dark web forums, and expert assessments from Dakota Cary (SentinelOne), Marc Hofer (NetAskari), and Jake Moore (ESET).

The authenticity of the dataset has not been independently verified. Chinese authorities — including the Ministry of Science and Technology and the Cyberspace Administration of China — had not issued a public response at the time of publication. The FlamingChina Telegram post dated February 6, 2026, and the associated dark web listing, remained the primary public sources of the claims. Further background on technology infrastructure shifts and platform-level changes in 2026 is available on Giganectar.

Leave a comment