Tech
Researcher plants 2m canaries into the AI supply chain. A privacy disaster surfaced.
Publication: Ars Technica
Author: Dan Goodin
Source published: 20/11/2026, 20:59:08 UTC
Briefing created: 20/11/2026, 21:14:20 UTC
Based on a feed excerpt; full article content has not been acquired. Source evidence retrieved 20/11/2026, 21:13:23 UTC.
Briefings
Researchers HD Moore deployed 2 million digital tripwires across AI services to detect data leaks. The experiment revealed a massive ecosystem where sensitive information leaked like a sieve through AI routers that select models based on cost rather than security. Moore used AWS credentials tied to empty accounts to seed secrets into prompts, allowing him to trace which specific AI services and providers were responsible for the exposure.
Key points
- Researchers HD Moore planted 2 million digital tripwires across AI services to detect data leaks.
- Sensitive information leaked through AI routers that select models based on cost rather than security.
Context
Moore's experiment tested if sensitive information leaks through AI services and how it is exploited. He used AWS credentials tied to empty accounts to seed secrets into prompts, allowing him to trace which specific AI services and providers were responsible for the exposure.
Limitations
Evidence is limited to Moore's specific experiment using AWS credentials tied to empty accounts. The study does not cover broader industry-wide security trends, long-term leak patterns, or the full scope of data types involved in the AI ecosystem.
Generated by Bittrees-hosted qwen35-2b sha256:aaf42c8b7c3c on 20/11/2026, 21:14:20 UTC. Based on publisher-provided feed evidence; may not cover the full original article. Check the source for details.
View archived briefing on IPFS
bafybeih3r6aeseyo7cztmtvwgua7bmumos4xo32fhyyenmvwzhxn3usrp4