Home/Events/Researcher plants 2m canaries into the AI supply chain revealing data leaks from multiple providers

Researcher plants 2m canaries into the AI supply chain revealing data leaks from multiple providers

Confirmed
Confidence
90%
Impact: 80%
Updated 8m ago

Consensus Brief

A researcher conducted an experiment by sending 2 million prompts containing unique secrets through various AI services, uncovering significant data leaks and security breaches. The findings indicated that at least three AI providers were compromised, leading to unauthorized access to sensitive information and violations of privacy settings.

What Changed Since Last Update

8m ago

New corroborating source added: Ars Technica published an update on Fri, 20 No ("Researcher plants 2m canaries into the AI supply chain. A privacy disaster surfaced.").

Claim Ledger

4 claims tracked across sources

Confirmed Fact

At least three AI providers were compromised by attackers who siphoned prompts.

Confirmed Fact

Some providers did not follow end-user privacy settings like Zero Data Retention.

Confirmed Fact

Secrets sent to two hacked providers were quickly reused from the same rented server.

Confirmed Fact

A Chinese AI lab's systems fetched secret URLs from prompts marked as 'no data collection'.

Role-Based Impact Analysis

Source Timeline

30 sources corroborating