2 Million Digital Canaries Reveal Leaks Across AI Service Chain
A researcher planted 2 million digital tripwires in requests sent through a sprawling network of AI services to see what happened to sensitive information entrusted to them. What he found was an...

A researcher planted 2 million digital tripwires in requests sent through a sprawling network of AI services to see what happened to sensitive information entrusted to them. What he found was an ecosystem that leaked highly sensitive data like a sieve.
The exposure surfaced through an experiment testing whether sensitive information can leak through the vast chain of widely used AI services and how that information is exploited when it does. Many of today's AI apps rely on services such as OpenRouter or Vercel that act as brokers. These "AI routers," sometimes called "AI gateways," receive large numbers of prompts from end users and route them to an available inference engine based on criteria such as lowest price or best fit for the task. Rather than using Anthropic, OpenAI, or another big provider, a developer hires an AI router to find the most suitable model.
To test for potential leakage, researcher HD Moore assembled a collection of AWS access keys, functioning credentials tied to empty AWS accounts, passwords for services such as SSH, 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 then seeded each secret into exactly one prompt sent to a single provider. If any secret leaked, these "canaries"-named for the birds once taken into coal mines-would let Moore identify each link in the chain responsible.
Source: Ars Technica
Ready to Modernize Your Business?
Get your AI automation roadmap in minutes, not months.
Analyze Your Workflows →