Home · Technology · Nov 20 archive
Researchers Uncover AI Supply Chain Leaks Using 2 Million Canaries
Confirmed
In Short: Ars Technica reports that researcher HD Moore conducted an experiment to test the security of AI services by planting 2 million digital canaries, or tripwires, into the requests sent through a network of AI services.

Supply chain network.
These canaries were designed to detect any leakage of sensitive information, such as AWS access keys, passwords, and unique URLs, that were intentionally included in the requests.
The canaries allowed Moore to trace the chain of responsibility if any secrets were leaked.
Marta Shahbazi, an award-winning researcher at the University of Cambridge, died in a Spanish train disaster while traveling to collect a prize for her work on human embryos and cell biology.
Researchers are also exploring the octopus as a model for natural contortionists, according to NPR, while other scientists are finding ways to connect seemingly unrelated items like plastic bottles and cookies.
What's confirmed
- These canaries were designed to detect any leakage of sensitive information, such as AWS access keys, passwords, and unique URLs, that were intentionally included in the requests.
- The canaries allowed Moore to trace the chain of responsibility if any secrets were leaked.
- Marta Shahbazi, an award-winning researcher at the University of Cambridge, died in a Spanish train disaster while traveling to collect a prize for her work on human embryos and cell biology.
- Researchers are also exploring the octopus as a model for natural contortionists, according to NPR, while other scientists are finding ways to connect seemingly unrelated items like plastic bottles and cookies.
What's still developing
- The exposure was found through an ingenious experiment designed to test if sensitive information can leak through the vast chain of widely used AI services and how that information is exploited when it does.
- Researchers planted 2 million digital tripwires in requests sent through a sprawling network of AI services to find out what happened to sensitive information entrusted to them.
- In the event any of the secrets were leaked, these “canaries”—so called because they resemble the role of that bird when taken into coal mines—would allow Moore to identify each link in the chain that was responsible.
- 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.
- The researchers presented their results at the fall meeting of the American Chemical Society (ACS) during the "Undergraduate and Graduate Research in Biochemistry and Chemical Biology" symposium at McCormick Place.
