Incident #1723
Anthropic's Claude Code AI Coding Tool Reportedly Used Hidden Unicode Markers to Classify Custom API Routing
Description
Researchers found that Anthropic's Claude Code AI coding tool classified custom API endpoint hostnames and two system timezones, encoding the results in system prompts with visually similar Unicode apostrophes and altered date separators. Anthropic said the mechanism was an anti-reseller and anti-distillation experiment launched in March and removed it after disclosure.
Actors
Alleged Deployers
- anthropic
- ai-service-providers
Alleged Developers
- anthropic
- ai-software-developers
Harmed Parties
- ai-coding-assistant-users
- privacy
- claude-code-users