The Pneumetron
NEWSTOOLSINFRASTRUCTURE
Pneumetron.AI supply chain test exposes credential theft across inference providers
Share
Skip to article content
  1. Home
  2. ›
  3. News
  4. ›
  5. Technology
  6. ›
  7. AI supply chain test exposes credential theft across inference providers
Technology·November 22, 2026

AI supply chain test exposes credential theft across inference providers

BY PNEUMETRON|4 MIN READ · 787 WORDS4 MIN READ
Tools
Share

In This Article

  • The canary test and what it found
  • The intermediary routing crisis
  • Privacy settings and data retention failures
  • Sources
Illustrative photo: switch, it, server, server cabinet
Illustrative image: blickpixel / Pixabay

Researchers planted 2 million canary tokens in AI requests, revealing that intermediaries routinely harvest credentials and ignore privacy settings. OpenRouter revoked retention labels for 125 endpoints after the findings.

Key Takeaways

  • 01Dragos researchers planted 2 million canary tokens in AI requests, finding at least three providers actively harvesting credentials.
  • 02OpenRouter removed three providers and revoked Zero Data Retention labels for 125 endpoints following the research.
  • 03Attackers tested stolen AWS keys within hours of harvesting them from routed prompts containing fake credentials.
  • 04Researchers noted the routing pattern matches known Chinese proxy services but could not confirm shared personnel.

Security researchers at Dragos planted 2 million canary tokens in AI inference requests, uncovering a supply chain where credentials and prompts are routinely harvested by third-party intermediaries. The researchers ran the tests from July through November 2026 and identified at least three compromised providers where attackers were actively stealing user data. None of the three knew they had been compromised until Dragos notified them, according to reports published by The Decoder and Ars Technica.

The canary test and what it found

To test for potential leakage, Moore assembled a collection of Amazon Web Services access keys tied to empty accounts, SSH passwords, and unique URLs, DNS names, and private git repository references. None of the secrets had ever been used before, and he controlled the infrastructure that received any requests using the credentials. He seeded each secret into exactly one prompt sent to a single provider. In the event any of the secrets were leaked, these “canaries” would allow the team to identify each link in the chain responsible for the breach. For developers building AI applications, the results mean that routing prompts through third-party brokers exposes sensitive data to any provider in the pool, regardless of the privacy guarantees advertised by the primary platform.

The results showed a system leaking sensitive information at scale. In the largest incident, attackers harvested hundreds of AWS keys, including some from requests the provider had rejected before processing. The attackers tested the stolen keys within hours, searched S3 storage, and attempted to establish persistent access. One attacker even used ARTEX, an AI-powered attack tool.

Marketplace resellers did not need to exploit software vulnerabilities to access this data. Simply listing a model at the lowest price and turning on logging was enough to capture customer requests. In one case, an individual on a residential internet connection in France used SSH passwords pulled from intercepted requests to attempt logins.

The intermediary routing crisis

The exposure highlights the risks of AI routers and gateways, which act as brokers between end users and inference engines. Instead of calling Anthropic or OpenAI directly, developers hire routers to find the most suitable model based on criteria like price or task fit. Requests are routed through a pool of providers, sometimes across several hops, and each hop can read the entire request and response.

In this case, providers including Novita, AtlasCloud, and StreamLake sat between OpenRouter and model developer MiniMax. Together runs its own infrastructure, and Phala has its roots in Web3. Following the research, OpenRouter responded by removing three providers from its platform and revoking Zero Data Retention approval for 125 endpoints. The company also emailed affected users, warning that third parties may have read requests sent to MiniMax models. The exposure involved prompts routed through resellers AtlasCloud, Novita, and StreamLake.

Illustrative photo: photographer, girl, camera, digital Illustrative image: geralt / Pixabay

Affected providers have published preliminary security reports on the incident. Dragos warns companies against sending requests containing credentials to intermediaries, recommending that organizations control how their AI requests are routed and plant their own canary tokens.

"None of this means that you should stop using AI services. It does mean that where your prompts go is a decision, and if you don’t make it, the lowest price will make it for you," HD Moore writes.

A recent analysis by the Oxford China Policy Lab showed how Chinese proxy services route requests to Western AI models through servers abroad. Those services can also read the requests in transit. It remains unclear whether the providers Dragos examined share any personnel with those proxy networks. The methods, however, follow the same basic pattern: intermediaries sit between users and model developers, view requests, and pass data to third parties without user knowledge. Data ends up with foreign actors even when users never intended to send it there, and privacy promises break down along the supply chain.

Privacy settings and data retention failures

The research also exposed widespread failures in data retention policies. Privacy settings like Zero Data Retention were frequently ignored. Dragos reported that a Chinese AI lab accessed secret test URLs in more than half of all requests, even when those requests went through US providers.

Even major US providers broke their own policies. One provider was still crawling a test URL more than two months after a single request. In another case, a secret test link appeared in an internal chat on Feishu (known outside China as Lark) 24 days after the request. The link was opened within two minutes on three computers and three phones in Beijing and Shanghai.

The findings suggest that developers relying on third-party routing to cut costs are effectively outsourcing data access to any provider in the pool, regardless of the privacy guarantees advertised by the primary platform.

Sources

  • the-decoder.com: Security researchers show how cheap AI inference providers steal keys and snoop on users
  • arstechnica.com: Researcher plants 2m canaries into the AI supply chain. A privacy disaster surfaced.
  • techcrunch.com: Meta’s AI agent Muse is mixing up conversations and could be exposing private data, report says | TechCrunch
#ai security#inference providers#supply chain#canary tokens#privacy
PE
PUBLISHED BY

PNEUMETRON EDITORIAL

Pneumetron's pipeline pairs AI-assisted drafting with human editorial review before publishing — our goal is to make staying informed easier, not to replace real reporting.

Source Material:the-decoder.com, arstechnica.com, techcrunch.com ↗
Share this article
Share
Stay Informed

Never miss a signal.

Subscribe to the Pneumetron Intelligence Digest.

← Previous
Nvidia Halts H200 Sales to China Amid Megaspeed Probe and Leadership Change

More from Technology

View All →
TechnologyJust now
Illustrative photo: network, server, system, infrastructure

Nvidia Halts H200 Sales to China Amid Megaspeed Probe and Leadership Change

Nvidia appoints interim CEO Colette Kress and halts H200 exports to China following a board investigation into Megaspeed's $4.8 billion chip purchases and alleged intelligence ties.

BY PNEUMETRON4 MIN READ
Read more
Technology12h ago
Illustrative photo: server, space, the server room, dark

OpenAI cuts reasoning extraction success to 0.2 percent; Signal blocks mobile screenshots

OpenAI reports adaptive extraction attack success rates dropped to 0.2 percent on GPT-6.2 Astra following new API filters and model training. Signal restricts screenshots on Android and iOS to prevent OS-level AI agents from reading private chats, citing incomplete privacy guarantees.

BY PNEUMETRON3 MIN READ
Read more
Technology12h ago
Illustrative photo: server, space, the server room, dark

Nvidia Reports $115.8 Billion Q3 Revenue Amid Vera Rubin Ramp And China Export Halt

Nvidia’s third-quarter revenue doubled to $115.8 billion, driven by a 110 percent jump in data center sales and a rapid Vera Rubin platform rollout. The company will halt H200 chip sales in China while navigating a leadership transition and an active board investigation.

BY PNEUMETRON3 MIN READ
Read more
Technology12h ago
Illustrative photo: server, cloud, development, business

Google rolls out tool retrieval, desktop automation, and telephony updates

Google is deploying three distinct agent infrastructure updates: an API optimization for large tool libraries, background desktop automation on Mac, and a telephony stack built on Gemini 4.

BY PNEUMETRON3 MIN READ
Read more

Comments

In This Article

  • The canary test and what it found
  • The intermediary routing crisis
  • Privacy settings and data retention failures
  • Sources

Most Read

01
Technology·3d ago
Nvidia Delays 10-Q Filing Amid SEC Probe and New AI Chip Export Controls
02
Technology·3d ago
Nvidia and AMD report record data center revenue as AI monitoring degrades
03
Technology·2d ago
Anthropic prices IPO at $83, opens trading at $111.50 amid safety disclosures
04
Technology·4d ago
Isomorphic Labs Secures $5B Series C to Advance AI Drug Design Platform
05
Technology·5d ago
Jensen Huang resigns as Nvidia CEO amid export scrutiny and open-source backlash
Daily Digest

Get top AI & tech signals delivered to your inbox every morning.

Subscribe →
Follow Signals
X / TWITTERLINKEDININSTAGRAMYOUTUBETELEGRAM

THE PNEUMETRON

The publication of modern technology, AI, and infrastructure.

AI-assisted, human-reviewed.

NEWS CATEGORIES

  • AI RESEARCH
  • TECHNOLOGY
  • WORLD
  • BUSINESS
  • POLITICS
  • SCIENCE
  • HEALTH
  • SPORTS
  • ENTERTAINMENT

AI & RESEARCH

  • AI RESEARCH
  • INFRASTRUCTURE
  • TOOLS
  • DEEP DIVES
  • TECHNICAL DISPATCHES

RESOURCES

  • ABOUT US
  • CONTACT
  • CAREERS
  • ADVERTISE
Pneumetron.
© 2026 Pneumetron. AI-assisted, human-reviewed.
All rights reserved.
ABOUTTOOLSPRIVACYTERMSCONTACT